Loanwords in Japanese
Studies in Language Companion Series (SLCS) This series has been established as a companion series to the periodical
Studies in Language. For an overview of all books published in this series, please see
http://benjamins.com/catalog!slcs Editors Werner Abraham
Elly van Gelderen
University ofVienna
Arizona State University
Editorial Boru:d Bernard Comrie Ma.x Planck Institute, Leipzig and University of California, Santa Barbara William Croft University of New Mexico Osten Dahl University of Stockholm Gerrit }. Dimmendaal University of Cologne Ekkehard Konig Free University of Berlin
Volume 125 Loanwords in Japanese by Mark Irwin
Christian Lehmann University of Erfurt Brian MacWhinney Carnegie-Mellon University Marianne Mithun University of Califomia, Santa Barbara Heiko Narrog Tohuku University Johanna L. Wood University of Aarhus
Loanwords in Japanese !\.1ark Irwin Yamagata University
John Benjamins Publishing Company Amsterdam/ Philadelphia
;"'""""''\ TM
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences Permanence of Paper for Printed Library Materials, ANSI z39.48--1984.
IOQ\
~ ..7
Library of Congress Cataloging-in-Publication Data
Irwin, Mark. Loanwords in Japanese I Mark Irwin. p. em. (Studies in Language Companion Series, ISSN 0165·-7763; v. 125) Includes bibliographical references and index. 1.
Japanese language-- Foreign words and phrases. I. Title.
PL664.A3L79
2011
495.6'5--dc22 ISBN
978 90 272 0592 6 (Hb; alk. paper)
ISBN
978 90 272 8689 5 (Eb)
2011009243
© 2011- John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without \-\Titten permission from the publisher.
John Benjamins Publishing Co.· P.O. Box 36224 · 1020 ME Amsterdam-The Netherlands John Benjamins North America· P.O. Box 27519 ·Philadelphia PA 19118-0519 · usA
1b K.&K.
Table of contents
List of tables
X1
List of figures
XIII
List of abbreviations
XV
Transcriptions & translations
XVII
Acknowledgements
XIX
CHAPTER 1
Introduction Language contact and lexical borrowing 1 1.2 Vocabulary strata in Japanese: Definitions & divisions 1.2.1 Native stratum 5 1.2.2 Mimetic stratum 5 1.2.3 Sino-Japanese stratum 6 1.2.4 Foreign stratum 7 1.2.5 Hybrid stratum 14 1.3 Vocabulary strata in Japanese: Distribution 14 1.4 Organization of this volume 21
1
1.1
4
CHAPTER 2
A history of Japanese loanwords 2.1 Overview 23 2.2 Iberian borrowings: Mid-16th century to mid-17th century 2.3 Dutch borrowings: Mid-17th to mid-19th century 35 2.4 \Vestern borrowings: 19th century to present 42 2.4.1 Russian borrowings 44 2.4.2 French borrowings 47 2.4.3 German borrowings ;o 2.4.4 English Borrowings 53 2.5 East Asian borrowings 61 2.6 Names 67
23
29
vm Loanwords in Japanese
CHAPTER 3
Phonology Phonemicization 71 3.2 Adaptation 76 3.2.1 Phonic substitution 81 3.2.1.1 Consonant'> 81 3.2.1.2 Vowels 94 3.2.2 Epenthesis 105 3.2.2.1 Vowel epenthesis 106 3.2.2.2 Mora obstruent epenthesis 3.2.3 Deletion 123 3·3 Supra segmental issues 126 3·4 Mora-clipping 129
71
:·P
CHAPTER
112
4
Morphology, morphophonology and semantics 4.1 Loanwords and morphemes, loanwords as morphemes 4.2 Morphological reduction 141 4·3 Compound phenomena 143 4.:P Compound reduction 143 4.3.1.1 Compound clipping 144 4.3.1.2 Ellipsis 148 4.3.1.3 Portmanteau f()rmation 149 4.3.2 Sequential voicing 150 4·4 Semantics 153 CHAPTER
137 137
5
Orthography 5.1 Japanese scripts 159 5.2 Loanword orthography until1955 169 5·3 Loanword orthography from 1955 173 5·3-1 Orthographical recommendations I: Monbush6 173 5.3.2 Orthographical recommendations II: Bunkach6 180 5·4 .Roman script 185 5-4-1 Abbreviations and full words 186 5.4.2 Acronyms 188 CHAPTER
159
6
Attitudes to Loanwords
193
Table of contents
References
207
Donor word index
229
J..oanword index
241
Subject index
271
IX
List of tables Table Table Table Table
1.1 1.2 1.3 1.4
Table 1.5 Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5 Table 2.6 Table 2.7 Table 2.8 Table 2.9 Table 2.10 Table 2.11 Table 2.12 Table 2.13 Table 3.1 Table Table Table Table Table
3.2 3.3 3.4 3.5 3.6
Table 3.7 Table 3.8
Thomason's (2001) borrowing scale (adapted and abridged) 2 Gaikokugo or gairaigo? 12 Hybrid word combinations 14 Distribution of Japanese lexemes by vocabulary stratum and frequency 17 Major analyses of vocabulary stratum distribution 19 Proportion of gairaigo by donor language (%) 27 Top five most frequent gairaigo by donor language based on the NINJAL (2006a) magazine corpus 29 Selection of early Catholic gairaigo from Portuguese and Latin 33 Selection of early secular gairaigo from Portuguese and Spanish 35 Selection of medical and scientific gairaigo from Dutch 40 Selection of mercantile and other gairaigo from Dutch 42 Selection of Russian gairaigo 47 Selection of French gairaigo 50 Selection of German gairaigo 53 Twenty most frequent English gairaigo in magazines 6o 'l\venty most frequent English gairaigo in the Mainichi Shinbun 1994-2003 61 Selection of Chinese gairaigo 65 Selection of Korean gairaigo 68 Permissible mora types in conservative and contemporary phonemic systems 75 Summary of consonant substitutions 95 Sources of gairaigo consonant phonemes 96 Adaptation of English rho tic vowels 102 Adaptation of English rhotic vowels: A broader synthesis 104 /Qf.-epenthesis in loanwords from English across three major environments u:~ Percentage of /Q/-epenthesis by English donor obstruent in WORD-FINAL environment 115 /Q/-epenthesis in loanwords from English: summary by environment and manner of articulation 119
xn Loanwords in Japanese Table 3.9 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 6.1 Table 6.2 Table 6.3
Table 6.4
Gairaigo truncation processes 130 The traditional kana 162 The contemporary katakana series 164 v-row transcription patterns across three media 183 Orthographic practice fcJr the 20 most frellUentgairaigo in magazines 187 Selection of Roman acronyms in Japanese 189 Top 5 most recognised, understood and used loanwords in BBK (2008) 194 Recommended replacements for the 20 least comprehended loanwords considered by the Loanword Committee 203 Comparison of length between the 20 least comprehended loanwords and the Loanword Committee's recommended replacements 204 Attitude oflocal government workers to Loan Committee replacement proposals 205
List of figures Figure 1.1 Figure 1.2 Figure 1.3 Figure L4 Figure 2.1 Hgure 3.1 Figure 3.2 Hgure 3.3 Figure 3.4 Figure 5.1 Figure 5.2 Figure 5.3 Figure 6.1 Figure 6.2 Figure 6.3
Japanese vocabulary stmta through time 4 Distribution of Japanese lexemes by vocabulary stratum: % token count 16 Distribution of Japanese lexemes by vocabulary stratum: % type count 18 Proportion of gaimigo tokens by media cluster: 1906···2006 2.0 Gairaigo timeline: 1500-present 26 'Ihe Japanese consonants: a typical analysis of a conservative phonemic system 72 The Japanese consonants: an analysis of the contemporary phonemic system 73 Borrowing routes in Japanese 77 Summary of VO'Nel substitutions 96 Typical Japanese newspaper article containing the three Japanese scripts 160 Conventional script usage in modern Japanese 167 'The use of Roman script in modern Japanese 186 Opinions on using loanwords in speech and writing 196 Reasons for a desirnble outlook towards loanwords 196 Reasons for an undesirable outlook tmvards loanwords 197
List of abbreviations
Ar. Arm. Bei. Cant. Cz. Dan. Du. Eng. Far. Fin. Fr. Gae. Ger. Haw.
Arabic Armenian Beijing Chinese Cantonese Czech Danish Dutch English Farsi Finnish French Gaelic German Hawaiian
He b. Hu. It. Kor. Lat. Nor. Pol. Por. Ru. Serb. Sp. Sw. Wel.
Hebrew Hungarian Italian Korean Latin Norwegian Polish Portuguese Russian Serbian Spanish Swedish Welsh
Transcriptions & translations
Japanese words cited as examples in this volume are transcribed according to Figure 3.2 on p. 73, the most salient features of which are that s represents an alveolo-palatal fricative, as in si, ·fitf ( L, ), 'poem'; can alveolo-palatal affricate, as in ci, :riJL ('S ), 'blood'; C a dental affricate, as in CU, )f. (~.J), 'harbour'; and Q the mora obstruent, conventionally written v) or ·~~,as in niQki, Ei ~G U<: ·~) ~),'diary'. Words not cited as examples, as well as names and titles, are given in Hepburn. Donor words from Beijing Chinese are transcribed in pinyin, from Cantonese in Yale, from Korean in McCune-Reischauer; donor Russian employs a maximal phonemic representation that indicates neutralization (e.g. word·· final devoicing) but ignores accenLrelated surface phonetics (e.g. vowel reduction); donor Ara·· bic, Farsi and Hebrew employ a broad IPA transcription; donor Ainu follows the romanization of the Saru dialect (Tamura 1996). For ease of reference, all Japanese linguistic terminology and Japanese per· sonal names have kanji (Chinese characters) appended the first time they appear in the text. All direct citations from original Japanese texts are the author's own translation.
Acknowledgements This volume is the product of a great deal of labour and much help has been received along the way: I owe an especial debt of gratitude to John Phillips, Mark Brierley, Nic Tranter, Heiko Narrog and Jeremy Williams, all of whom painstakingly read through earlier drafts and provided many comments and corrections. Wearing his cartographical hat, Mark Brierley also kindly drew up the map on page 24. Many others have helped in countless ways, large and small. In no particular order, I would like to thank Nakamura Tadashi and Aizawa Naoki for help with Russian; Hong Ja-Eul for help with Korean; Yu Ming Huang and Sophie Wang for help with Chinese; Morita Mitsuhiro, Suzuki T()ru and Watanabe Fumio for help with Japanese; Jimmy Lam for help with Dutch; Sakai Keiko and Sasaki Yuki for cross-checking attestations; as well as Todd Enslen, Tim Greer and Graham Healey. It goes, of course, without saying that all errors and inadequacies are entirely my O'Nn. Yamagata, autumn 2010
CHAPTER 1
Introduction
1.1
Language contact and lexical borrowing
language contact drives a range of linguistic phenomena, among them pidginization, language mixture, language shift, code-switching, bi- and multilingualism, grammaticalization,language death and borrowing (Thomason & Kaufman 1988; Holm 1988, 1989; Romaine 1989; Winf(Hd 2003). This volume concerns itself with the last of these phenomena, specifically lexical borrowing into Japanese over the past 450 years. The extent of lexical borrowing, and the level of contact which drives it, vary hugely from language to language. Table 1.1 is a summary of Thomason's (2001:70--71) 'borrowing scaiC, restricted to the lexicon and ignoring other types of borrowing such as 'sounds, phonological rules, grammatical morphemes, syntactic patterns, semantic associations, discourse strategies or whatever' (Campbell 2004:62). In terms of scale, the degree of contact f(mnd in Japanese falls into either the 'slightly more intense' or the 'more intense' category. TI1e borrowing of nouns, verbs and adjectives is commonplace. While there is very little evidence as yet for the borrowing of adverbial particles or of conjunctions (the infrequent aNdooa from Eng. ami/oris one of the few examples), we do find adpositions (oobaa and iN from Eng. over and in), personal pronouns (mai and yuu from Eng. my and you), basic vocabulary such as low numerals (waN and ii, both 'onC, from Eng. one and Bei. yi), derivational affixes (furu and izumu from Eng. -Jul and -ism) and even articles (za and ra, from Eng. the and Fr. la). These borrowings are rarely used as tunction words in Japanese but rather as nouns, verbs or affixes (§4.1). Recent years, though, have witnessed an increase in phrases where they are used as function words, especially in advertising: e.g. izakaya iN nagasaki 'pubs in Nagasaki', written in mixed script (iN in the Roman alphabet, the remainder in Japanese scripts). The '(slightly) more intense' lexical borrowing evident in Japanese has parallels in, for example, the history of English or 'Ihrkish. What is significant about lexical borrowing in Japanese, however, is that there has been little direct linguistic contact. In terms of borrowers, contact is 'casual' on Thomason's (2001) scale. Borrowing has occurred in a strongly monolingual societ}' and within the physical confines of
2
Loanwords in Japanese
Table 1.1. Thomason's (2001) borrowing scale (adapted and abridged) Degree of contact
Scal.e of lexical borrowing
Borrowers
casual
mostly twuns, some verbs, adjectives and adverbs; non-basic vombulal)'' nouns verbs adiectives adverbs
no fluency itl source language required; few bilinguals
slightly more intense
more intense
intense
conju~ction~ adverbial'particle~
reasonably fluent bilinguals forming minorif:v of borrowing language speakers more bilinguals, social factors favour borrowing
non-basic vocabulary nouns, verbs, adjectives, adverbs, conjunctions, adverbial particles, derimtional affixes and more; basic and non-basic vocabulary heavy borrowing ofail vocabulary extensive bilirlgttalism, social factors stmng~v favour borrowing
a state with few speakers of other indigenous languages.l1hat Japan is an ethnically homogeneous nation has been widely discredited: see, f()r example, the papers in Weiner (1997a). However, a debunked myth of Japanese ethnic homogeneity should not be cont1ated with a debunked myth of Japanese linguistic homogeneity - that Japan is not ethnically and culturally homogenous does not mean Japan is a multilingual land overflowing with bilinguals. Combined speaker numbers for all non-indigenous languages likely do not exceed 1% of the population2 and the overwhelming majority of the inhabitants of Japan today, whatever their ethnidty and whatever their culture, have, to use Bloomfield's (1933:56) definition ofbilingual-ism, 'native--like control' over only one language: Japanese. Even by definitions of
1. Grimes (2000) lists fifteen living languages for Japan, including Japanese, Japanese sign language and the non -indigenous Korean. Of the remaining twelve, eleven are languages of the Ryukyuan group. Together, their total speaker numbers are less than 1 million (approximately 0.8% of the population of Japan), communities during the last century having gone 'from a predominantly monolingual Okinawan language speaking society to a state of bilingualism without diglossia, a stage alongthewayto becoming monolingual inJapanese'(Osumi 2001:73). According to Matsumori (1995: 39), there are probably no longer any monolingual speakers of the Ryukyuan languages remaining (see also Heinrich 2004). The one remaining indigenous language, Ainu, is dying. On Hokkaido in 1993 it had 'no more than ten or twenty native speakers: according to Vovin (1993: 1), and 'less than ted, according to Janhunen & Salminen (1993-1999). Hattori (1967: 58) states that Ainu 'has reached the point of complete extinction: Shibata (1990:3) that it is 'on the brink of extinction: See also DeChicchis (1995: 109-lll). 2. Although detailed statistics on language use are unavailable, speaker numbers can be inferred from government statistics such as STK (2008) and HNKK (2009). The most widely used non-indigenous languages found on the archipelago today are probably Chinese topolects and Korean, followed by Portuguese, Tagalog, English and Spanish.
Chapter 1. Introduction
bilingualism looser than Bloomfield's (e.g. Mackey 1968), the bilingual proportion of the Japanese population has been very small during the four and a half centuries covered in this volume (see Maher (2001) f(lr discussion of possible bilingualism on the Japanese archipelago in earlier historical periods). Language contact in Japan since the 16th century has thus taken place for the most part in a distant setting. Thomason (2001: 1---2) defines language contact in its simplest form as 'the use of more than one language in the same place at the same time, but later notes that this definition is 'implicitly flawed ... : speakers of two (or more) languages need not be in the same place for language contact to occur'. Further, the motivation behind distant language contact in Japan has not been religion, commonly f(mnd in other paradigms, and one motivation behind earlier borrowing from Chinese (§1.2.3). Instead, it has been the prestige accorded Western science and philosophy (§2.3, §2.4), and the languages, in particular English, in which such knowledge is disseminated. Hand in hand with this has gone the huge emphasis placed on comprehension of the written rather than the spoken Western word. The adaptation patterns shown by most Japanese loanwords reflect this distant, orthographic contact (§3.2), are an important illustration of Thomason's 'implicit flaw: and make loanwords in Japanese an essential area of study. Few general works on Japanese loanwords have been published. The earliest and by far the most authoritative general work is Umegaki (1963). This is an updated second edition of a first edition published 20 years previously containing research carried out mainly in the 1930s. It is out of print and has become somewhat dated. Ishiwata (2001) is more recent and focused heavily on etymology and semantics. Other, less wide-ranging, general works include Ishino (1983), Tanaka (2002) and Kobayashi (2009), as well as Jinnouchi (2007), who analyses loanwords from a sociolinguistic perspective. All these volumes are in Japanese and thus inaccessible to the general English reader. The sole general work published in a Western language is Loveday (1996). Extreme caution is urged Miller's (1998) review of the book is scathing, yet tair. Short sections on loanwords appear in many Japanese general works on the Japanese language too many to list here. The more important will be referenced during the course of this volume. In Western languages, such works are necessarily fewer. Worthy of mention are the relevant sections in Miller (1967), Vance (1987), Shibatani (1990), Gottlieb (2005) andLabrune (2006). Scholarlyarticleson)apanese loanwords begin, in Japanese, ·with Otsuki (1884), then Maeda (1922) and Arakawa (1931); in English, with Ichikawa (1929, 1930); and, by non--native speakers of Japanese, with Daniels (1948) and Carr (1951). Research in recent decades has been focused strongly on phonology, morphophonology and morphology, especially mora obstruent epenthesis (see §3.2.2.2 for a list of previous research), accent (Sibata 1994; Katayama 1995; Shinohara 2000; Kubozono 200lb, 2002a,
3
4
Loanwords in Japanese
2006; Kubozono & Fukui 2006) and truncation processes (It6 1990; It6 & Mester 1992; Kuwamoto 1998a; Kubozono 1999b, 2001a, 2002a; Labrune 2002). Finally, a programme of mainly survey-based research carried out by the Kokuritsu Kokugo Kenkyujo (National Institute for Japanese Language, hereafter NINJAL) has been continuing for nearly 60 years and is a crucial source of diachronic data. Particularly important have been NINJAL (1962, 1964, 1984, 1987, 2005a).
1.2
Vocabulary strata in Japanese: Definitions & divisions
Japanese exhibits lexical stratification. Native (§1.2.1), mimetic (§1.2.2), SinoJapanese (§1.2.3) and foreign strata (§1.2.4) are all well established, with most scholars acknowledging the existence of three or four of these layers. 3 The four vocabulary strata to be outlined below are schematicized diachronically in Figure 1.1, the jagged line around the mid-8th century indicating the limit of written evidence. A hybrid stratum is discussed briefly in §1.2.5.
Native
Mimetic
SinoJapanese
Foreign
SOOCE
1000CE
1500CE
Present
Figure 1.1. Japanese vocabulary strata through time
3· For example, Martin (1952), Backhouse (1993) and Gottlieb (2005) posit three layers, while McCawley (1968), Vance (1987), Shibatani (1990) and Nishio (2002) posit four_ A few scholars (e_g_ Rice 1997; Ota 2004) are sceptical as to the existence of lexical stratification in Japanese at all. Rather than models based on lexical strata, It5 & Mester (1995, 1999), as well as Fukuzawa, Kitahara & Ota (1998), Fukuzawa & Kitahara (2005) and others, have posited core-periphery or set-inclusion models (2, p. 9).
Chapter I. Introduction
1.2.1
Native stratum
Words belonging to the native Japanese stratum are known in the Japanese literature as wago 5ftlil1k or yamatokotoba 7c5f'P -~~.The former designation literally means 'words of wa: where wa ;,fP is the modern Japanese fc)rm of an ancient Chinese term ~ (Bei. wo) for a culture or cultures located on the Japanese archi·· pelago. It is used in modern Japanese as a prefix meaning 'Japanese: 'Ihe word wa itself is therefore, somewhat ironically, not a member of the native vocabulary stratum, but of the Sino-Japanese, and appears in numerous Sino-Japanese compounds such as wa.Soku 'Japanese cuisine' and waeijiteN 'Japanese-English dictionary'. Yamatokotoba are generally glossed in the English literature as 'Yamato words' or 'Yamato vocabulary: Yamato being the ancient name for a clan, a province and then a ruling court centred on the modern prefecture ofNara, but which by the early 20th century had taken on strong nationalistic overtones and had come to refer to the dominant ethnic group within the Japanese Empire. Japanese academic usage favours the less emotive wago and it is this term, as well as 'native: which I shall employ in this book. As is apparent from these Japanese terms and their English translations, the native layer is composed of those words which have not been borrowed from other languages and are 'indigenous' to Japanese. Cross-linguistically, it is well known that 'basic' or 'core' vocabulary is the most resistant, though not immune, to borrowing 4 and it therefore comes as no surprise that a large proportion of Japanese core vocabulary is wago. These include grammatical particles such as ya 'and', wa 'topic marker', dake 'only'; fundamental biological activities such as naku 'crY, taberu 'eaf, Sinu 'die'; kinship terms such as haha 'mother: musuko 'son', mago 'grandchild'; body parts such as atama 'head: te 'arm, hand: hiza. 'knee'; and natural phenomena such as yuki 'snow', nami 'wave: cuki 'moon'. Nevertheless, some areas of the Japanese core lexicon are non-native. Numerals greater than ten have been replaced wholesale by Sino-Japanese, are obsolete unbound and survive bound only, if at all, in compounds and names. 1.2.2
Mimetic stratum
Words belonging to the mimetic (or ideophonic or phonaesthetic) stratum are known in Japanese asgise1go jtji~- or giongo if: it~. This stratum may be further subdivided, although categorizations vary from scholar to scholar and a detailed analysis is beyond the scope of this introduction. Shibatani (1990: 153-157), for
4 e.g. Bynon (1977:256) or Hock (1991:385-386). See Beckwith (2004: 195-213) for a more critical view.
5
6
Loanwords in Japanese
example, offers a tripartite division into phonomimes (giseigo), phenomimes (gitaigo t'fr~ [if) and psychomimes (g~jogol' ;it [if). Closest to the conventional sense of onomatopoeia are phonomimes which mimic sounds (gacaN 'crasH, zaazaa 'the sound of pouring rain'), while phenomimes mimic the state or manner of external phenomena (pikapika 'shinY, caracara 'cheap and flashy, tacky') and psychomimes mimic psychological conditions (zukizuki 'throbbing: bikubiku 'scared'). Whichever subcategorization one adheres to, mimetic vocabulary is charac-terized by reduplication and a number of phonological and syntactic constraints. As Hamano (1998: 11) notes, \·\'hile 'the majority... are never entered into ordinary Japanese dictionaries or textbooks' and 'even advanced students of Japanese fall short of fully comprehending [their] nuanceS: f(n native speakers 'their expressive meanings are readily understood'. No mimetic words have been borrowed and thus there is a strong claim for subsuming the mimetic stratum within the native, as do many scholars who posit only three lexical strata. 1.2.3
Sino-Japanese stratum
Words belonging to the Sino-Japanese stratum are generally knm·vn in Japanese as kango :1jUg- 'Han words' or, less hequently, as }iongo ::fo~:.~ 'words made from the sounds of Chinese characters'..Han refers to the .Han Dynasty (206 BCE 220 CE) from 'Nhich the modern ethnonym Han, the dominant ethnic group of the Peo-ple's Republic of China, is derived. Most kango are, as the name suggests, words borrowed from Chinese. Borrowing from China was intense and occurred over a period of approximately one and a half millennia, beginning with the adoption of Chinese characters in the Japanese archipelago, for which the earliest evidence dates from approximately the 5th century CE (§5.1). Not all kango were borrowed directly from Chinese, however. A significant number, knm·\'n collectively in the Japanese tradition as wasei no kango :fq~i_(1)jjtill} 'kango made in Japan, are words created hom the readings attached to Chinese characters (§2.5). Wasei no kango consist of two main classes. A medieval class first appearing c. 1000 and continuing for three or four centuries, where the Chinese characters used to \·vrite some wago words came to be read as kango (Shirafuji 2005: 211; Chin 2007); and a second loan translation class (known as kango yakugo ~t"iif ~~. ~g 'kango calques' or shinkango ;fr-)~Mf 'new kango') which were the product of 'Dutch learning' (rangaku ~- ~; §2.3) and the later pursuit of rapid industrial and economic growth by the Japanese government from the mid--19th century (§2.4). Here, Sino-- Japanese played the same role as Greek and Latin in the creation of scientific vocabulary in many European languages during the Renaissance. Sino-Japanese words reflect the immense prestige accorded Chinese culture throughout the centuries: Buddhist terminology (often originally from Sanskrit), such as kyoo 'sutra' and sooryo 'monk'; Confucian notions, such as kookoo 'filial
Chapter I. Introduction
piety' and seNpai 'senior, elder'; political and administrative terms, such as daijiN 'minister' and koomu 'public office, public duties'; astrological terms, such as juuni.~i 't
a. information b. thug c. chutzpah
horrible kangaroo zeitgeist
transparent tomato sang-froid
Any educated English speaker who has learned even a modicum of French but who has no background in linguistics \\
a
1.2.4
.Foreign stratum
Words belonging to the foreign stratum are known in Japanese as gairaigo :7r·:>lt ~ ('foreign words'), shakuyogo 1~ JfJ ~ ('borrO\·ved words'), yoga J?f.*" ('Western words'), or katakanago J.t'LR~ ~-('words written in katakana: the script in which they are now ty11ically written: see §5.1.). Yogo is used only of loanwords borrowed from European languages and thus excludes borrowings from modern Chinese, Korean and Ainu (§2.5), while katakanago is a recent, more colloquial term. Of the two remaining Japanese terms, shakuyogo has a broader meaning and may also encompass the Sino--Japanese stratum (see discussion in Tanaka 2002: 32···35). I will use gairaigo throughout this volume.
7
8
Loanwords in Japanese
Gairaigo may be grossly defined as the residue after native, Sino-Japanese and mimetic words have been removed from the lexicon. But as a definition this is insufficient. Gairaigo are defined by the prestigious Japanese dictionary, Nihon Kokugo Daijiten (NKD), as words taken from one language into another and u<>ed as part of the recipient language. In Japan, this signifies words that have entered Japanese since the end of the Muromachi period [1573] fi·om countries outside the Sinosphere. Howeve1; words borrowed from Chinese in the modern era (e.g. ii ~')ne: gyooza 'steamed dumpling: kooryaN 'type of millet') may also be considered gairaigo. Moreover, there are gairaigo such as oorudornisu 'old maid, spinster' (from *old miss) and naitaa 'night game (in baseball)' (from •·nighter) 5 which are not used in the donor language ... Words borrowed via Chinese but ultimately deriving from Sanskrit (e.g. neha,v 'nirvana; ajari 'Buddhist master') are not usually considered to be gairaigo and are often written in Chinese characters.
This definition raises some interesting issues. A fe\-\' Portuguese gairaigo (§2.2) are attested in writing pre··1573 (Table 2.3) and it is likely that some attested post-·1573 were in fact borw
He then goes on to divide gairaigo into three broad groups: Chinese, European and 'other'. Leaving aside his unfortunate identification of'country' with 'language,
5· The English nighter, although no longer used in this sense, did in fact exist around 1950. See Ueno (1980:107-115).
Chapter I. Introduction
Sugimoto's definition is pleasingly sweeping. His definition of 'other' encompasses extremely ancient loans from Korean, as well as borrowings whose ultimate source is Sanskrit. All of these were borrowed well bef(xe any notional Muromachi cut-·off date and all fall outside NKD's definition. On the other hand, Sugimoto defines all borrowings from Chinese as gairaigo: not just very ancient borrowings, such as uma 'horse' and ume 'plum: but modern Chinese borro'~Nings and, crucially, the huge quantity of words that make up the Sino--Japanese stratum. Sugimoto's defi-nition is, nonetheless, thoroughly logical: it simply takes in anything that is not a native Japanese \·\'ord. A theoretical approach to what constitutes gairaigo can be found in It0 & Mester's (1999) core-periphery model of vocabulary stratification, reproduced here in modified form: (2)
Lexmax f--
Lexmax
I
f--
Lex2 - Lex1
assimilated foreign
I
f--
Lex1 - Lex0
established loans
f--
Lex0
Lex 2
unassimilated foreign
Lex2 Lex 1 Lex0 ....
native
The model's structure is organized in terms of set inclusion, so that members of the innermost core (native Japanese words) 'fulfil lexical constraints in the maximal way: while lexical items in outlying strata 'fulfil only a small subset of. .. constraints'. This set inclusion structure is, they claim, the lexi·· con's fundamental characteristic, the 'existence of large, homogenous and welldefined strata' being 'a secondary phenomenon'. '!heir four sets, lex0 - Lexmax, are based on four constraints and their systematic violations within the phono·· logical lexicon as a whole. In the context of the present discussion, the two rel-· evant constraints are No-DD (no voiced obstruent geminates: see §3.2.2.2) for 'assimilated foreign' and SYLLSTRUC (the basic syllable constraints of Japanese, including no complex onsets or codas: see §3.1) for 'unassimilated foreign'. 'Assimilated foreign' thus differ from 'established loans' (Sino Japanese) and 'native' vocabulary by virtue of permitting voiced obstruent geminates, although they still suffer from syllable constraints, from \·\'hich 'unassimilated foreign' do not. Loveday states that [s]trictly linguistic judgments concerning the way an item is pronounced or its faithfulness to the donor meaning are not sufficient to determine its gairaigo
9
10
Loanwords in Japanese status. The main determining factors for recognizing gairaigo would seem to be: the degree of the item's historic assimilation, and its general level of intelligibility in the community, both of which reveal themselves in the extent of its orthographic. and phonetic. conformity to Japanese norms and in its increased occurrence. Loveday (1996:49-50)
While I disagree with Loveday's contention that orthographic conformity is a determining factor, I concur with intelligibility and phonetic conformity. However, his definition offers no criteria for distinguishing Sino--Japanese from gairaigo. Given this multiplicity of views (and I have omitted several others), how then should gairaigo be defined? For the purposes of this book I adopt the following definition: (3)
A gairaigo is a foreign word. which has undergone adaptation to Japanese phonology, has been borrowed into Japanese after the mid-16th century and whose meaning is, or has been, intelligible to the general speech community.
I posit no specific date before which a word cannot be a gairaigo, but rather the broad 'mid-16th century: corresponding to the first Japanese contact with European languages (§2.2). Under my definition, gairaigo include Chinese and Korean words borrowed since this time but exclude all borrowings prior, i.e. early Korean and Chinese loans as well as some others, such as raQko 'sea otter' from Ainu rakko. Loans ultimately from Sanskrit are excluded for the same rea-son. Any word borrowed from Chinese before the mid--16th century, along with all kango yakugo (§1.2.3, '§2.4), I consider Sino-Japanese, any borrowed thereafter gairaigo. Under my definition, a gairaigo must have undergone adaptation ('§3.2), while orthography plays no role. I return now to the distinction between (1b) and (1c) above. As noted, the examples in (lb) have been adapted to English phonology, while those examples in (lc) are very recent borrowings and may not have been. In the Japanese literature, words such as (lb) correspond to gairaigo, while those cited in (lc) are generally referred to as gaikokugo :1f· ~ lif.· or 'foreign words'. Gaikokugo not only fail to undergo adaptation, thus corresponding to Ito & Mester's (1999) \massimilated foreign' set in (2), they are also typically unintelligible to the general speech community. For gaikokugo I offer the following definition: (4)
A gaikokugo is a foreign word which has not undergone adaptation, or a foreign word whose meaning has always been unintelligible to the general speech community.6
6. l do not claim that agairaigo-gaikokugo distinction ought to be drawn between (lb) and (I c) in English, nor that, if one were, it shm:ild be based on my definition in (4). I have used
Chapter I. Introduction
The issue of adaptation will be described in detail in Chapter 3 and requires no further discussion at this juncture. The issue of intelligibility is somewhat nebu·· lous, and since I do not treat gaikokugo to any great extent in this volume, is an important factor in determining their exclusion. Table 1.2 attempts to summarize the ensuing discussion. No visitor to Japan can fail to notice the multitude of f.r>reign \W>rds written in the Roman alphabet, overwhelmingly English. These are found in advertising (f, o), on packaging (n) and clothes (j), in pop song lyrics (g) and band names (k), in magazine titles (e), in television programme and film titles (h, 1), in company names and in brand names (j, m). They may be misspelt (m) and may contain grammatical mistakes (n) or oddities (k, o) (Gabbrielli 2005; Jinnouchi 2007: 59-75). With film titles, in particular, the meaning is often obscure or unintelligible to the average Japanese, perhaps a deliberate ploy on the part of marketing departments to lure viewers into watching material which would be ignored were its content obvious from its title (Yamada 2001, 2005: 172ft} Although most of these foreign words are adapted when produced orally, some (n) are rarely spoken orally but simply read or 'vie\·\'ed: being essentially decorative and analysable as 'monophonic part \-\'riting' (Backhaus 2007: 90···103). However, not all words such as these are gaikokugo, if the notion of intel·· ligibility, as defined in (3) and (4), is taken into account. For example, (e), the title of a magazine marketed at teenage girls, although irrelevant to its content, is intelligible. Typically written in native rather than Roman script, poteto 'potato' has not only undergone adaptation but is widely understood both by the general speech community and the target readership of the magazine in question. Falling one month after Valentine's Day, (f) is a semantically remodelled compound (§4.4), where two words borrowed from English have been assembled into a compound whose meaning, though different from any a native speaker of English would extract, is nevertheless widely understood among Japanese speakers. On the other hand, the meanings of the brand name (i), band name (k) and film title (l) are unintelligible to the general Japanese speech community and, although all have been adapted, are gaikokugo. A donor--targeted pronunciation such as (p), although typically intelligible due to the existence of (a), is not adapted and, there·· fore, also gaikokugo. Even more clearly gaikokugo is (q): here gaikokugo takes on its broader meaning of'fi)reign language'.
(I) for purposes of analogy only. Similarly, my gairaigo-gaikokugo distinction differs from the traditional division found in German linguistics between Lehnwort and .Fremdwort (see the definitions in Dudenreaktion 1993-95).
n
Utterance
b
Status
Orthography
irisuto/
noun
l) ,'"2.., l-·
/~1Qkusui
noun
FAX, fitx, 7"'1' -y-.,"'1,'"2.., .BMW 'IU.tel,
Script
/poteto/
magazine title semantically remodelled' compound
POTATO ;;j•;TJ_,.( }-'f'-
flyby
foreign
p-¥·;:t:t-
native
:L~]-:'7
native
pop lyrics extract from songby(k)
h
/roodoobuzariNgu/
t1lm title
/epidora/
+f· 1l>'7
ye.s ye.s
widely widely
gairaigo gairaigo
Eng. potato
yes
widely
ga.ira~'So
March 14th (2·17, p.156)
Eng. white,
yes
widely
ga.ira~'So
flyby
Eng.J1yby
yes
specialist usage
Lordofthf
~ng.Lordof
yes
partially
G~r;iEn~.
yes
specialist usage
Rings
j
/faato/
shirt brand
FART
foreign
k
ibaNpuobutikiNi
pop group
foreign
/teNjiriNgu/
t1lm title
B1Jl\1P OF CIDCKEN 5"• .::c. /':?)1) :/J}'
i8'"reedo/ /jereedo/
hair gel brand
GERAlD
foreign
n
/foameNheva wairudisur~imuj
strapline on (m) packaging
FOR MEN HI\IR WIIDSLLTI;ffi
foreign
0
/iNoupaiazanekisuto/
Hita
INSPIRE TilE I-<"EXT
to reign
p
[list]
donor-targeted pronundation of(a)
list, ') :J, 1-
foreign
m
Eng.
native
e f
/furaibai/
Ger. BJ.t'f},r
telephone number potato
typkally foreign foreign
abbreviation
g
BMW
f01eign
acronym
/tere/
Jhowaitodee/
widely widely
ga.ira~'So
yes
list fax
ibiiemudaburyuu/
7·v
Eng. list Eng. fax
native typicallv flneign'
d
native
epidural
Source
tele(phone)
Intelligible?
guikokugo
day
gairr
tneR1.ngs
epr.aura,. Eng.fart Eng. bump, ofi chicken Eng. charrgeli,-eg
yes
ga.ikok:u.go ga.ikok:u.go
yes
ga.ikok:u.go
Eng.~el?,
yes
ga.ikok:u.go
Eng.}?t' men hai;~ wild. slime
yes
ga.ikok:u.go
inopiretl1e next generation?
fh~~;r_;~·?i.;
yes
8aiko'
list
Eng.ii
Changeling
yes
Ger. el?) Eng. a.id?
or native
Gairaigo or
Phonotactio:allr adapted. yes
Meaning
probably
8aiko'
Chapter I. Introduction
It is not always the case that a word \·\'hich has been adapted to Japanese is \'\'ritten in native Japanese orthography (c), for example, is rarely if ever found written in native script, but instead appears in Roman letters. As it is spoken according to an adaptation of the English (occasionally the German) letter names - B is bii, M is emu and W is daburyuu- and is widely intelligible, I consider itgairaigo and not gaikokugo. Similarly, (b) and (d) are frequently found \Hitten on everything from posters stuck on lampposts to advertisements in the daily press. Both are found only infrequently in native script, (d) especially so (Table 5.4), but when they are produced orally, they undergo adaptation. These, too, are gaimigo. Conversely, although the film title (1) is written in native script, it is unintelligible to the general speech community and is thus gaikokugo. Grey areas nevertheless exist and awkward questions remain unresolved. The film title (h) is adapted but only partially unintelligible: za and riNgu are not uncommon, obu is occasionally encountered, but modo with the meaning 'lord' is obscure. To the average Japanese speaker unaware of the fact that the film was given a different title from the book (translated into native Japanese as yubiwa monogotari 'Tale of the Ring') or unaware even of the existence of the book, this phrase's status as gairaigo or gaikokugo is moot. Although specialist terms such as (g) and (i) are intelligible, they are only so to a restricted group of users, in this case pilots and medical staff. The gairaigo or gaikokugo status of words like these is also open to question. What proportion of a population must fail to comprehend a borrowing for it to be unintelligible to the general speech community? If the vast majority of a particular socio·economic group, age cohort, ethnic group, gen .. der, dialect group, religion or profession comprehend a borrowing, but the vast majority of the remainder of the Japanese-speaking population do not, should the borrowing be considered gairaigo or gaikokugo? How well-known must a person, place or country be .f(n a name to be classified as a gairaigo? What constitutes intelligibility? What constitutes the general speech community? Ultimately, de fin·· ing the boundaries between gairaigo and gaikokugo is highly problematic and I make no claims as to the universality of the definition oftered here. It must serve, nevertheless, for the purposes of this volume. The oldest gairaigo date back to borrowings introduced by Portuguese· and Spanish-speaking missionaries in the latter half of the 16th century (§2.2); these were followed by a two-century science-driven wave of borrowings from Dutch between the late 17th and the mid .. J9th centuries (§2.3); and then, propelled by ever increasing internationalization, succeeded by borrowings from other European languages (§2.4). 'Ioday, although the overwhelming majority of foreign borrowings are from English (Table 2.1), many of these suffer from varying degrees of unintelligibility, with the most acute levels of incomprehension
13
14
Loanwords in Japanese
apparent amongst the elderly (Table 6.2). Such borrowings thus stand on the cusp between gairaigo and gaikokugo. Hybrid stratum
1.2.5
Finally, there exists the hybrid stratum, whose members are known in the Japanese tradition as konshugo ~It1'.it~. If one acknowledges the existence of all the other f(mr vocabulary strata (native, mimetic, Sino-Japanese and foreign) and restricts examples to compounds with two elements, then there are 12 possible hybrid com·· binations, as shown in Table 1.3. Of these combinations, those with mimetic elements are the least frequent, and those having a mimetic as their second element the least frequent of all. Table 1.3. Hybrid word combinations native native .Sjr.K!::l.~P..~H~§.~
mimetic
foreign
1.3
Sino-Ia12anese
mimetic
foreign
slka,J.Uk!J 'venison'
maruvoca 'plunip' k..iNpika 'gilded'
wagomu 'rubber band'
kimagure 'whim' poisute 'littering'
dotabataki!i[eki 'slapstick comedy'
koohiimame 'coffee bean~
eQkususeN 'X-ray'
r.uggw.~~J,meeru
'spam (email)' biQkurimaaku 'exdamation mark'
camu.5abu 'lamb shabushabu'
Vocabulary strata in Japanese: Distribution
Over the past century; across a range of media, a number of surveys have been carried out whose results reveal the proportions of the various strata comprising the Japanese lexicon. The two largest and most comprehensive surveys examined the vocabulary of a broad··based magazine sample and were carried out by NINJAL. The magazines sampled were published in 1956 and 1994 and the survey data published as NINJAL (1962, 1964) and NINJAL (2005a, 2006a),7
7· NlNJAL (2006a) is the electronic version ofNINJAL (2005a) and is available online in a number of formats, including easily manipulable Excel spreadsheets. Although the raw data are split between a large jiritsugo \$1 .:dif 'independent word' file and a smallfuzokugo MJ~ jlj- 'dependent word' file, I have amalgamated the two data sets and arne nded frequency
Chapter I. Introduction
respectively. Pigure 1.2 shows vocabulary stratum distribution in these two surveys by token count: i.e. the total number of words. In the four decades between the two surveys, Sino--Japanese has overtaken native vocabulary as the largest stratum. What is perhaps more conspicuous is that the proportion of foreign lexemes has quadrupled from 3% to 12%. Just under one-eighth of all tokens appearing in the pages of the 1994 magazine sample were gairaigo. Figure 1.3 shows vocabulary stratum distribution in the same two surveys, this time by type count, i.e. the total number of different words. By this criterion, while foreign lexemes accounted for nearly one-tenth of all types even in 1956, by 1994 this figure had risen to over one--third. Although now the dominant stratum by type count, most individual gairaigo nevertheless have low frequency rates. This is illustrated in Table 1.4. Only 5 types have a token count higher than 25,000 in the circa 700,000-word NINJAL (2006a) corpus: all are native lexemes. Gairaigo do not signify until the 500-999 frequency band, of which they make up 8%. Thereafter, the proportion of gairaigo rises steadily, but does not overtake native/mimetic vocabulary until the 5---9 frequency band and only becomes the dominant stratum (37%) amongst the least frequent lexemes, the 1-4 frequency band. Native/mimetic lexemes show the opposite trend, with their proportion falling as token counts decrease. Sino--Japanese lexemes show a bimodal distribution, peaking once in the 5000---9999 frequency band and again in the 50-99 band. Over the decades, NINJAL has carried out other smaller vocabulary surveys from which data on vocabulary stratum distribution can be extracted. NINJAL (1952) surveyed Asahi Shinbun newspaper editions published in June 1949, while NINJAL (1970) analysed newspapers published in 1966. Gairaigo made up 5% and 11% of types, respectively, in these two surveys. NINJAL (1983b) and (1986) examined senior and junior high school textbooks used in the 1974 and 1980 academic years, loanwords comprising 2% of tokens in both surveys. NINJAL (1987) is a corpus based on 17 magazines published at 10-year intervals between 1906 and 1976: see below for further detail. Vocabulary studies published by NINJAL staff include Tanaka (2006), who surveyed the electronic version of the Mainichi Shinbun and a selection of government white papers and local government information leaflets; and Yamaguchi (2007) who analysed the contents of the electronic version of the Mainichi Shinbun from 1994 to 2002.
rates for the purposes of Table 1.4, as well as Tables 2.2, 2.10 and 5.4. All calculations based on the NINJAL (2006a) raw data here, and elsewhere in this volume, are my own and exclude names. The author is grateful to NINJAL for putting such data in the public domain.
15
16
Loanwords in Japanese
In Tanaka gairaigo accounted for 3% to 4% of tokens, while in Yamaguchi the figure was 5%.
1956: o/o Token count •
3% 2%
Native/mimetic
D Sino-Japanese •
Foreign
D Hybrid
54%
1994: o/o Token count •
Native/mimetic
D Sino-Japanese •
Foreign
D Hybrid
Figure 1.2. Distribution of Japanese lexemes by vocabulary stratum:% token count (NINJAL 1964, 2005a)
Other studies of the written word relevant to vocabulary stratum distribution cover a range of different media. In an analysis of books aimed at elementary school age children (6-11 year-olds), Nomura & Yanase (1979, 1989) found that gairaigo accounted for 4% of types, with non-fiction evincing higher levels than fiction. Satake (1991) looked at the writing of young Japanese across four media: 'youth magazine' letter columns, an art gallery comments box, fan letters to an author of books aimed at the young, and a random survey of extracts from a widerange of 'junior novels'. Here the first category showed the highest proportion of gairaigo (5% of tokens), with the second category showing the lowest (1 %). The unweighted average across all four categories was 3%. High levels of gairaigo were evinced in a study by Ito (2003) on the lyrics of the complete works (320 songs) of the pop singer Matsutoya Yumi ;f't-1.£-~ 1E ~-Here, gairaigo made up 15% of types and 5% of tokens. As noted in §1.2.4, Japanese pop lyrics contain a significant number of gaikokugo. These accounted for a further 8% of types and 10% of tokens
Chapter 1. Introduction Table 1.4. Distribution of Japanese lexemes by vocabulary stratum and frequency (extracted from NIN.JAL 2006a) %SinoJapanese
% 11oreign
%Hybrid
100%
0%
0%
0%
88%
13%
0%
0%
13
38%
62%
0%
0%
2500-4999
24
46%
54%
0%
0%
1000-2499 500-999
54 71
59% 52%
37% 38%
0% 8%
4% 1%
250-499
181
51%
39%
8%
2%
100---249
613
37%
47%
14%
2%
Number of types
%Nati.ve/ mimetic
25000+
5
10000-24999
8
5000-9999
1"requency (token count)
50-99
993
37%
47%
15%
1%
25-49
1,557
29%
45%
23%
3%
3,736
27%
45%
26%
2%
·'1,987 32,189
26%
42% 30%
29%
4% 8%
10-25 5--9 1-4
25%
37%
in Ito's study. An even larger proportion ofgaimigo was found in Takashi's (1990a) small study of magazine and newspaper advertisements. Here, some 16% of types were loanwords. Yamada (2005: 270-330) examined the number of gairaigo per page across 362 novels published between 1871 and 2003. Novels analysed were restricted to one per author and averaged just under three per year. The propor-tion of gairaigo in this medium saw a significant increase, with types increasing approximately fourfold and tokens approximately ninefold between the 63 novels surveyed in the period 1871 1911 and the 62 novels surveyed in the period 1989---2003. Hashimoto (2007) looked at the number ofgairaigo in Asahi Shinbun and Yomiuri Shinbun editorials over the last htmdred years and found a 20- to 30-fold increase in tokens. The steepest increase occurred between approximately 1940 and 1970. In surveys of the spoken word, Shibu & Sanada (1980) analysed the speech of 25 people working or attending university and found that foreign vocabulary composed 10% of types and 3% of tokens. The level of gairaigo found in NINJA:Ls (1999) analysis of television programmes varied significantly according to programme genre: from as high as 16% for sports programmes to as low as 6% in educational programmes. Gairaigo occur with even greater frequency in television commercials: 23% of types according to Takashi's (1990a) small-scale study.
17
18
Loanwords in Japanese
1956: o/o Type count
•
Native/mimetic
•
Sino-Japanese
•
Foreign
D Hybrid
1994: o/o Type count
6%
•
Native/mimetic
•
Sino-Japanese
•
Foreign
D Hybrid
Figure 1.3. Distribution ofJapanese lexemes by vocabulary stratum: %type count (NINJAL 1964, 2005a)
Finally, dictionary-based studies include Miyajima's (1958) analyses of Genkai (Otsuki 1889) and Reikai Kokugo ]iten (Tokieda 1956), Saneto's (1973) analysis of Kadokawa Kokugo ]iten (Hisamatsu & Sat6 1969), Nomura's (1984) analysis of two editions of the specialist dictionary Gendai Yoga no Kiso Chishiki (JKS 1960, 1980) and Kindaichi's (2002) analysis of Shinsen Kokugo ]iten (Kindaichi 2002). For more detail on these surveys, see Table 1.5. This Table lists, in chronological order, the proportion of gairaigo in all the surveys just reviewed, with the exception of Yamada (2005) and Hashimoto (2007), where results are not given as percentages. Any comparison between these various surveys which ignored the range of different media across which they have been conducted would be problematic. Figure 1.4 thus groups into nine media clusters the 22 studies in Table 1.5 which provide token data, then plots the proportion of gairaigo vocabulary tokens across time. The 'magazine' and 'newspaper' clusters require no further explanation. The 'school textbook' cluster comprises NINJAL (1983b) and NINJAL (1986), the 'spoken word' cluster Shibu & Sanada (1980) and NINJAL (1999). Other clusters contain material carried out in the same calendar year (the two
survev
carri~dout. 1889 1906 1916 1926 1936 1946 1949 1956 1956 1956 1960 1966 1966 1969 1973--2002 1974 1976 1977-1978 1978' 1980 1980 1988--1989 1989 1989 1989--1990
1994 1994-2002 2002 2003 2003 2003
Token Miyajima (1958: 24) NINJAL (1987: 57) NIN)AL (1987: 57) NIN)AL (1987: 57) NINJAL ( 1987: 57) NINJAL (1987: 57) NINJAL (1952: 104) Miyajima (1958: 24) NINJAL (1987: 57) NIN)AL (1964: 61) Nomura (1984: 42-46) NINJAL ( 1987: 57) NINJAL(1970: 17) Saneto (1973: 343-344) Ito (2003) NINJAL (1983b: 43) NIN)AL (1987: 57) Shibu & Sanada (1980: 72) Nomum &Yanase (1979, 1989) NINJAL (1986: 33) Nomura (1984: 42-46) Takashi (1990a: 149) Takashi (1990a: 149) NIN)AL (1999: 82) Satake (1991: 7) · NINJ AL ( 2005a:32) Yamaguchi (2007) Kindaichi (2002) Tanaka (2006: 43--46) Tanaka (2006: 43-46) Tanaka (2006: 43-46)
dictionary headwords (39,103) 1 magazines (c. 10,000) magazines (c. 10,000) magazines (c:. 10,000) magazines ( c:. 10,000) magazines (c. 10,000) newspapers ( 14.,4.1 9) 0 standard dictionary(40,393) 0 magazines (c. 10,000) magazines (•Ul,972) speciali&t dictionary (9,910) magazines ( c:. 10,000) newspapers(40,589) standard dictionary((l0,218) pop song lyrics0 senior high school textbooks (321,058) magazines (c. 10,000) university student and employee speech (66,239) books for 6·-11 age group (14,081) junior high school textbooks (250,572) specialist dictionary (23,448) magazine and newspaper advertisements (2,610) television commercials (843) television programmes (103,081) yow1gpeople'swriting" , newspaper letter column s< magazines ( 693,173) newspapers (c. 315,000,000)d standard dictionary (86,775) 0 government white papers (10,000) public information leaflets (10,000) newspapers (10,000)
0.4' 0.4' 0.6' 0.8' 0.7' 1.4' 2.9 1.7' 4.81 1.8 2.3' 3.2 2.1 1.9
4.2 3.0 1.9 12.3' 5.0
Type 1.4 0.8' 0.7' 1.2' 1.4' 1.2' 4.7 3.5 2.4' 9.8f 43.1 fg 2.9' 11.2 7.8 14.Sf 6.1 3.9' 10.1 4.0 5.0 57.61B
15.7 23.3 12.4 ()
34.8f 8.6
8.8 2.8 3.4 3.9
' books in print, not necessarily published, in the year in question; 0 type cow1t; ., neither token nor type count stated; 0 morpheme token count, including numerals, punctuation and other symbols;' includes as gaimigo words ultimately from ~~n~kt"1t_ hnt t3Yrlncl~ nbrt> n~mpo;::;• f PvdncJp.;::; n;lnlP:'\:• g P:Yrlnilt>~ Rom::m ::Jinh~hPt ~cronvnH: ::~nd ;lhhrt3v;~tinn~ {fS.f:\_4.\
::r .§
.., ('O
a,.., 0
p..
~ ......
o·t:l
20
Loanwords in Japanese
'government publication' surveys), or composed of a single survey ('pop lyrics: '6-11 year-old books: 'young people's writing: 'newspaper letter columns').
14,-------------------------------12+---------------------------~---
+
10+--------------------------------
D Newspapers
Magazines
..t.. School textbooks
8+-------------------------------6+--------------------------------
1926
1946
1966
1986
X Pop lyrics )K
6-11 y.o. books
e
Spoken word
+
Young people's writing
2006
Figure 1.4. Proportion of gairaigo tokens by media cluster: 1906-2006
Despite being clustered by media type, caution is still required. The methodologies of the various studies differ substantially; names, acronyms and abbreviations may or may not be included; and definitions of both 'word' and 'gairaigo' vary. Even so, two things are immediately apparent from Figure 1.4. Firstly; the inexorable rise in the proportion of gairaigo vocabulary across most media since the first survey providing token data was carried out in 1906. This rise appears to accelerate slightly in the 1950s and then significantly from the 1980s. Secondly, the seemingly aberrant nature of the figure of 12% returned by the survey carried out on magazines published in 1994 (NINJAL 2005a, Figure 1.2). This may be explained in two ways. Either the medium of magazines has shown an increase in gairaigo penetration far steeper than that recorded in other media, or there was a fundamental difference between the sampling methodology and/or statistical analysis involved in NINJAL (2005a) and those magazine surveys preceding it (NINJAL 1964, 1987). There is little evidence for the latter. All earlier magazine surveys by NINJAL include advertisements in their statistics, although no specific figures are given. NINJAL (2005a) states specifically that 21% of all tokens were extracted from advertisements. Other clusters in Figure 1.4 do not include advertising data, however. This is noteworthy, given Haarmann (1989) and others have demonstrated that advertisements show the highest level of gairaigo penetration of any written medium. It is probably safest to conclude that
Chapter I. Introduction
the medium of magazines has undergone an acceleration in gairaigo penetration, due to an increase in the proportion of gairaigo found in advertising, an increase in the proportion of advertising found in magazines, or both. As dictionary surveys consider only type counts, these have not been included in Hgure 1.4. In any comparison between dictionaries, as with other media, caution is required..Methodologies and definitions vary and editorial policies differ. Nomura's (1984) analyses of Gendai Yogo no Kiso Chishiki (JKS 1960, 1980) show especially high levels of gairaigo, not surprising given these dictionaries focus on recent technical terminology, much of \·\'hich has been borrowed.
1.4
Organization of this volume
This volume is organized as follows. In Chapter 2, I present a diachronic account ofgairaigo. Loanword sources can be divided chronologically into three: an initial Iberian phase driven by Catholic proselytizing, beginning in the mid-16th century; two centuries during which Dutch merchants \·vere the sole source of \Vestern scientific learning at a time when the Japanese shogunate espoused a policy of isolationism; and a modern international phase beginning with the opening up of Japan after the Meiji Restoration of 1868 and, since World War II, dominated by English borrm\'ings. Chapter 3 examines the phonology of gairaigo. Central to this is the question of ho'N donor sounds are adapted to the Japanese phonemic inventory. Three main strategies are apparent and these can be summarized by reference to the loanword sarabureQdo, borrowed from Eng. thoroughbred. Hrstly, phonic substitution has meant that the donor word-initial sound not part of the Japanese phonemic inventory; has been adapted to Japanese lsi; the donor vowels represented by (o) and (ou) have both been adapted to Japanese /a/; and the remaining sounds in
e,
21
22
Loanwords in Japanese
reduction commonly occurs during the borrowing process, as the many gairaigo shorn of their plural, tense and possessive morphemes attest. This chapter also examines in detail two gairaigo compound phenomena, sequential voicing and compound reduction, the latter encompassing compound clipping, ellipsis and portmanteau formation. During or after the borrowing process, semantic shift may occur, most notably narrowing and pejoration. Finally, semantically remodelled compounds are examined. These are composed of two or more independent gairaigo, borrowed mostly from English, which when compounded take on a meaning deviating from what the semantic outcome of an identical compound would be in the donor language. Examples such as beQdotauN 'commuter town' (from bed+ town) hold potential pitfalls for English language learning in Japan, while provoking amusement and causing comprehension difficulties for Englishspeakers studying Japanese. Chapter 5 considers perhaps the most characteristic aspect of the Japanese foreign stratum: its orthography. Although gairaigo are typically written today in a native script generally not used by other vocabulary strata, this has not always been the case. After a diachronic examination of pre-modern gairaigo orthographic practice, the 1955 and 1991 recommendations of two government bodies are reviewed. The chapter closes with an examination of the increasing use of the Roman alphabet in writing gairaigo. The final chapter, Chapter 6, looks at attitudes to loanwords. After examining one of the many surveys carried out on levels of recognition, comprehension and frequency of individual loanwords, the results of a range of mostly governmentsponsored surveys into public attitudes to loanwords are dissected. Also investigated are the attitudes ofbureaucrats and civil servants, a sector ofJapanese society often accused of loanword overuse. The volume closes with an assessment of the recently formed Loanword Committee's proposals for loanword replacements and a review of public and academic reaction to them.
CHAPTER 2
A history of Japanese loanwords
2.1
Overview
Japanese loanword history can be divided into three broad streams: Iberian (§2.2), Dutch (§2.3) and Western (§2.4). The first phase runs from the mid-16th to the mid-17th century and was initiated by Portuguese-speaking Jesuits. Spanishspeaking missionaries followed, as did Portuguese and other foreign merchants. The bulk of the loans in this period are from Portuguese, though a small number from Latin, Spanish and Dutch are also found. The Dutch merchants permitted to remain in Japan after 1639 were the engine for the second loanword phase, during whose two centuries the Japanese absorbed contemporary European scientific learning. Running from the mid--19th century to the present day; the most recent 'Western' phase refers initially to the languages of the pre-World War I European Great Powers - Russian (§2A.1), French (§2.4.2), German (§2.4.3) and English (§2A.4) - but latterly, with the rise of American economic and political power, almost exclusively to American English. Figure 2.1 is a gross schematization of the level of borrowing from the main donor languages during these three phases. There has also been a small amount of borrowing from other ·western languages, as well as from modern Chinese topolects, Korean and Ainu (§2.5). The borrowing of place names and brand names is examined in §2.6. What is notable about Japanese loanwords is that their donor languages are typically geographically distant. While Chinese was almost the sole source of borrowing before gairaigo history begins (§1.2.3), it figures little thereafter. Other neighbouring languages, such as Korean, Ainu and Ryukyuan have also provided minimal input. Despite periodically sharing a land border with Japan, there has been relatively little borrowing hom Russian. Regional politics have, of course, been a major contributory factor here. Japan adopted a policy of sakoku, or quasi-isolation, from 1641 to 1859, during which ordinary Japanese were forbidden to leave the country and most foreigners refused entry. TI1e gradual collapse of the Qing dynasty in China, culminating in imperial overthrow, civil war and communist revolution in the first half of the 20th century, ushered in a corresponding weakening of Chinese cultural prestige. Relations with neighbouring Russia, and with the USSR, have rarely been good and are still poor. Finally,
24
Loanwords in Japanese
4
0
P (ETOROFU)
4
3
0
2
0
PACIFIC OCEAN
IMA
Chapter 2. A history ofJapanese loanwords
the rise of militarism in Japan from the early 1930s led to an imperialism which typically viewed the cultures, and languages, of its colonized lands on the Asian continent as inferior. While this imperialist phase led to a decade-long de facto eschewal of linguistic borrowing (5, p. 57), post-war the floodgates opened on gairaigo.
Heavy
, ............ __ _ Dutch
g>
Moderate
I I
0
..c "-
0
1'
Portuguese
-~
; ,, ..................
Slight
]
Russian
-····-····-····-·················
Marginal
1500
1600
1700
1800
1900
2000
Figure 2.1. Gairaigo timeline: 1500-present
What sets apart the gairaigo history to be described in this chapter from previous literature on the subject - almost all in Japanese - is that dates of first written attestations are provided. The two major sources of such attestations are NKD (2000-02) and Arakawa (1977). These are prefixed with 0 and •, respectively (e.g. 0 1888, •1888). Saito (1967) is a further source of written attestations, though for Dutch only, and these are prefixed with + Any attestation whose source is cited as being a dictionary is ignored- these I regard as potential gaikokugo (4, p. 10). Equally, I ignore any attestation where only a source, but no attested form, is cited. This is not uncommon with Arakawa. Naturally, the probability exists that earlier written attestations remain unrecorded by the dictionaries' editors. Over the years a number of methodologically diverse surveys have analysed the donor languages of gairaigo found in dictionaries, magazines and other materials. The results of these are listed in Table 2.1. Each survey's definition of gairaigo may differ from that in (3) on p. 10. None includes foreign place names. Ueno (1980: 78-81) is a survey of vocabulary appearing in the comic novel Aguranabe (Kanagaki 1871-72). Umegaki (1963:73) cites the figure for Englishgairaigo found
25
Period surveyed
;!I
1872 1889 1901-1920 1906 1916 c.1920 1921-1940 1926 1936 1941-1960 1946 1956 1956 1956 1961-1980 1966 1973-2002 1976 1981-2000
~
0 F
,
~
t
6
35
0
26
3
%Bytype
Source
31 Ueno (1980: 78-81) Umegaki (1963: 73) 551 Hashimoto (2006)" NINJAL (1987: 65)b 32 29 NINJAL (1987: 65)b 162 Ueno (1980: 104) Hashimoto (2006)" 49 NINJAL (1987: 65)b 63 NINJAL (1987: 65)b Hashimoto (2006)" NINJAL (1987: 65)b 50 NINJAL (1964: 64)" 2964 NINJAL (1987: 65)b 99 710 Ishiwata (1960)" Hashimoto (2006)" NINJAL (1987: 65)b 129 Ito (2003: 47)" 693 NINJAL (1987: 65)b 181 Hashimoto (2006)"
i 29
G)
~ 0
%Byto.ken
=
i
G)
~
~
0 F
,
0
~
=
~
i
li"' II>
13 6 17
0 3
3 7
13 7 7C
56 59 63
13 3 2
9 3 4
71 76
8 8
6 6
6 2
2 0
4 3
2 5
74 81 84 84
4 3 3 3
8 6 2 5
6 1 0 3
2 1 1 0
2 1 2 1
4 7C
87 85 83
4 1 1
4 1 5 2 5 2
1 2 2
2 1 0
0 1 1
2 5 7
1 4 . 13
6 5c
40 32
53 56
10 3
8 3
10 19
0 3
5 9
15 6
301 58 77 605 60
83 67 73 89 77
5 9 6 2 3
4 5 8 2 7
5 5 1 4 5
0 3 0 0 2
1 9 5 0 3
2 2 6 3 3
136 1558 2370 159 1447 220 3095
88 87 88 86 87 82 91
2 2 5 4 1 1 4
1 4 2 4 5 4 2
0 2 3 1 3 1 1
1 0 0 3 0 0 0
1 1 0 0 1 1
5 4C 2 2 4 10
Chapter 2. A history of Japanese loanwords
in the dictionary Genkai (Otsuki 1889). 8 Hashimoto (2006) is an analysis of the gairaigo appearing in .Asahi Shinbun editorials in 20--year blocks over a century. Ueno (1.980: 104) is based on a study by Maeda (1.922) of 200 selected 'frequent' gairaigo of the 'mid-Taisho Era'. Maeda's donor languages have been adjusted by Ueno to those cited by Kojien (Shinmura 1969), an authoritative Japanese dictionary: Since the gairaigo have been 'selected: the results of this study should be treated with caution. Ishiwata (1960) is a preliminary analysis of the gairaigo eventually published in NINJAL (1962, 1964) - the author was a member of the research group. 9 For a description of the NINJAL (1964, 1987) and Ito (2003) surveys, see §1.3 and Table 1.5. Although some of the earlier sample sizes are small and statistically unreliable, the data in Table 2.1 reflect the inexorable rise of English as the dominant global language and the 'total availability' (Miller 1967: 249) of the English lexicon for Japanese neologism. From a relatively small base of only 13% of types in 1.889 (I ignore the statistically insignificant 1872 survey), English loanwords comprised the majority of the gairaigo stratum by the turn of the 20th century. By around 1960 the proportion of English loans had levelled off at the 85% mark. Though no statistics for gairaigo donor languages by type are available after 1.976, it is probable the proportion of English loanwords has not changed substantially since this date. Until and unless English loans begin to oust already \-\'ell-established loans from other languages, an equilibrium of a sort appears to have been reached. Dutch and .Portuguese loans have seen the greatest drop in share, from 14% and 1.3% of types respectively around 1920 (Ueno 1980) to only a few percentage points today. None of the surveys listed in Table 2.1 include figures f()f recent Chinese loans. This reflects the high degree of academic neglect these borrowings have suffered and is an issue I take up in §2.5. An analysis of NINJAL (2006a), a large magazine corpus which includes text from advertisements (§1.3), provides another angle on gairaigo donor languages. Table 2.2 shows the five most frequent gairaigo borrowed from each of the six Ianguages covered in Table 2.1. The most frequent of any gairaigo in NINTAL (2006a) is mirimeetoru 'millimetre, borrowed from .French. Together with its truncated (§3.4) form miri, it has a frequency of .76102%o, or approximately one occurrence
8. The total count for all gairaigo in this dictionary is from .Miyajima (1958): see Table 1.5. I have not included in Table 2.1 the breakdown of gairaigo listed in the afterword of Genkai due to its equivocal categori7.ation (see also Hayashi et al. 1982: 71). 9· Donor languages in Ishiwata are based on etymologies cited in both the 1'-'leikai (Kindaichi 1943) and Kojieu (Shinmura 1955) dictionaries. In most cases, these are identical and the figures in Table 2.1 are an average of the two.
27
28
Loanwords in Japanese
in every 1,315 \·\'ords. This has been normalized to a figure of 100 and used as a benchmark against which other gairaigo are measured..From Table 2.2 it can be seen that loamw>rds borrowed from French account for the three most fre .. guent gairaigo and that the ten most frequent gairaigo are all from either hench or English. This level of penetration by h-ench may appear inconsistent with Table 2.1, but is readily explicable by the importance in Japanese everyday life of metric measurements, into which category all five of the French gairaigo fall. Compared to l~rench and English, loanwords from the other four major donor languages show considerably lower frequency rates. The most freguent German gairaigo is only one .. seventh as common, the most frequent Dutch one .. ninth as common, the most fi-equent Portuguese one-f()urteenth as common, and the most frequent Russian less than one-hundredth as common as mirimeetoru.
Table 2.2. Top five most frequent gairaigo by donor language based on the NINJA I.. (2006a) magazine corpus .~~~~ .......... ~~s.~~.~~ ......................................................~~.~~.~.~ .......................................................... J.~.~.~~~ ....................................................... 1. t'aipu 'type' S4 teema 'theme' 14 mi1"imeetotu > mi1·i" ' 100 seq to ·ser are.n.tgi'i ··allergy' 2 53 6 91 seNCimeetoru > -~NCi '<:m' t-errfoN > tere 'telephone' 53 st•t·oqku '(ski) stock~ teet~>n.t 'm' 4 58 4 5
saizu 'size'
Rank
Dutch
eJ~jiN 'engine'
46 .tJ
a.·ru.ba:ito > l,"IQ.iro (part-time joli uretaN curethanC
3 3
Portuguese
k1:rameeroru 'knf guramu'g Rus.•ian
52 51
··:···················reNi;~·cielti································Tr········s;;taN·ct;ut:t<;·-r-······································y-·······;~;;;c;ra·:·;~o~~kacv<;dkA;·············
4
gamsu 'glass' hiiru 'beer'
6 6
gasu'gas' eki-;u. extract'
5 5
paN'bread' ttlba.ko 'cigarette' teNpura 'tempura' jub,m'undershirt'
4 3
i,l;;·..,,·a 'salmon roe'
<1
noruma "quota'
2
iNten: •intellectual'
<1 <1 <1
peresuto·roika 'perestroika'
There are a number of gaimigo dictionaries on the market providing infor.. mation on loanword sources. The larger of these include Arakawa (1977), SSDH (1979), Ishiwata (1990) and Maruyama etal. (1992), as well as the specifically etymological gaimigo dictionary Yoshizawa & Ishiwata (1979). However, caution is warranted when using so-called katakana dictionaries for etymological research purposes. These tend to be smaller and slimmer and the use of the word katakana (the script in \·vhich gairaigo are typically written: §5.1) in their title hints at a mass-market audience. As Ishh·\'ata (2001: 170) notes, 'katakana dictionaries have gradually come close to being ordinary Japanese·English dictionaries, where their purpose is no longer to provide one with a loanword's etymology but with h
Chapter 2. A history of Japanese loanwords
guides- an English-language learner can quickly find how an English word is spelt through searching for its katakana adaptation.
2.2
Iberian borrowings: Mid-16th century to mid-17th century
What I term the 'Iberian periocf corresponds to what is known in the Japanese historical tradition as the nanbansen ~ ~~- period. Ncmban, literally 'southern barbariarl, \·\'as originally one of four derogatory designations used by the Chinese to refer to barbarian tribes living, for the most part, outside their empire. After the Portuguese and Spanish had established trading posts and colonies in south and southeast Asia, the term came to be applied by the Japanese to those foreigners \·vho approached their land from the south and who themselves hailed from southern Europe. Nanbansen means 'nanban ships' and refers not just to those in which Catholic missionaries arrived at the beginning of the period, but also to those in \·vhich sailed the Iberian merchants who followed in their \·vake. The gairaigo dating from this period, nearly all borrowed from Portuguese, the bulk of the remainder from Spanish or Latin, are divided into two groups in the Japanese literature: kirishitan, often tmnslated as 'Christian; and boeki 'trade'. Since the Christianity introduced to Japan at this time \·\'as solely that of the Roman Catholic Church - there were no Protestant, Orthodox or other missions - the term kirishitan is perhaps better translated as 'Catholic'. A.s not all 'trade' vocabu·· lary refers to tmdable goods some place names, for example, belong here the second group is better termed 'secular'. I turn first to the Catholic loans. Japan's first regular contact with the European \·\'orld began in 1543, when some Portuguese sailors, possibly including the adventurer--explorer Fernao Mendes Pinto, 10 landed on the small southern island of Tanegashima, part of the Osumi island group in \·\'hat is now Kagoshima Prefecture. Prompted by a meeting \·\'ith a fugitive Japanese murderer, Yajiro (also known as Anjiro, later as Paolo de Santa Fe), in .Malacca a year and a half previously, Francis de Xavier (1506-1552), Navarrese co--founder of the Society of Jesus, arrived in 1.549 in what is now Kagoshima City, thus marking the beginning of the Portuguese Jesuit missions to Japan. Spanish mendicant orders, including Dominicans and Franciscans, followed. The history of their preaching and conversions, the lives of their Japanese followers, and the establishment of Catholic churches and missions is tmced in detail in \W>rks by Boxer (1951), Cooper (1974) or Higashibaba (2001, 2006) and need not concern us
m
10. 'Ihe account of Pinto's various explorations, Peregrinm;am (Peregrinations), is regarded by scholars as generally unreliable (Boxer 1951: 18-27).
29
30
Loanwords in Japanese
further here. Subsequent political change within Japan does. The Tokugawa shogunate, which established itself in Edo in 1603, sought to clamp down on the regional autonomy of the daimyo, the feudal lords. Their resistance, as well as the Shimabara Rebellion (modern Nagasaki Prefecture) of 1637-38, was blamed on Christianity. Christians, foreign and Japanese, were persecuted. The Spanish missions were expelled in1624, the Portuguese missions in 1636. While physical contact was initially limited for the most part to Krushu (although by 1560 missionary schools had also been established in the Kinki region), and its conduit typically restricted to Catholic missionaries, Japan's first direct contact with European ideas, religion and everyday objects was a major watershed in its history. Its impact is apparent in the mass of borrowings dating from this period, not just from Portuguese, but also from Latin. The latter is still the language oflegal documents in the Catholic church today and one in which most educated Europeans of the time would have been competent. The majority of the Portuguese Jesuits and Spanish monks were from these educated classes. They brought with them objects not only from Europe but also from elsewhere in their missionary sphere (India, China, Korea, Java), many of which were also of great interest to the Japanese. Thus, a few loanwords from other Asian languages entered the language, including kiseru '(smoking) pipe' [ 0 1612] from Khmer. The missionaries undertook the task of compiling dictionaries, such as the 1603 Japanese--Portuguese dictionary Nippo Jisho 'Vocabulario da Lingoa de Iapam' (Ishizuka 1976; Doi 1980); 11 Japanese grammars, such as Nihon Bunt-en ~t\rte da Lingoa de Iapam' (Rodriguez 1604); and Roman alphabet transcriptions of Japanese literature, such as Feiqe no lvionogatari (Fabian [1592] 1967). These works, known as the kirishitan shiryo 'Catholic materials: are of huge importance fen our modern knowledge of the Late Middle Japanese period for, among other reasons, they include the first Japanese texts written in the Roman alphabet. Thorough treatments of this literature can be found in Doi (1963, 1971) or Ebisawa ( 1991). The Jesuits were aware from the beginning that the locals they were seeking to convert would not understand Portuguese or Latin religious terms. For this reason, they initially employed loan translations in an attempt to convey Catholic doctrine: e.g. dair1iCi for 'god: tamasii fi)r 'souf, teN for 'heaven' (Schurhammer 1928; Doi 1933). These attempts ended in failure, since many were Buddhist terms having unintended or negative meanings. The missionaries therefore made the decision to eschew translation, in consequence promoting borrowing. How many words were
n.
See Cooper (1976) for the question of authorship.
Chapter 2. A hlstory of Japanese loanwords
borrowed from Portuguese and other languages during this period is difficult to estimate. Regions of Japan in closest contact with the Catholic missions, Kyu.shu and especially modern Nagasaki Prefecture, preserved more words for a lengthier period. Even standard Japanese dictionaries list such dialect terms as baNko 'bench: from Por. banco, found in various Kyushf1 dialects. In his monumental dictionary of Nag-c:tsaki gairaigo compiled during the first half of the 20th century; Koga (2000) records some 3,245 words of Portuguese and 227 of Latin origin. Table 2.3 offers a taste of those words borrowed to describe elements of the Catholic faith- its liturgical procedures, hierarchical nomenclature, theology and doctrine. In the table, and throughout this chapter, many loanwords appear in more than one form in the historical record. Ishiwata (2001: 197 -198) notes, for example, that batereN appears as bateree, bateNreN, bootere and bateru. I list only the most commonly found form(s). In the table, and again throughout this chapter, I cite source words in their modern spelling, give their meaning only when this differs from that of the Japanese loan, and cite gairaigo in modern standard Japanese pronunciation. How 'intelligible to the general speech commw1ity' some of the words in Table 2.3 were is open to question. So, therefore, is their gairaigo status according to my definition in (3) on p. 10.12 Examples of the more widely comprehended loans include the word for Catholic converts themselves (and their associated practices), kiriSitaN, and, more conspicuously, what has now become the word. for bread in modern Japanese, paN. This had originally referred to the commu-· nion wafer (Umegak:i 1963:52), but later underwent semantic broadening (241, p. 154). By the turn of the 17th century, government suppression of Christianity had led to many Catholic gairaigo becoming taboo and by the time freedom of religion was granted in 1871, most had become obsolete outside extremely select circles. 13 As Miller (1967: 240) has put it, these loanwords 'became precious and
12. Matsuoka (1982:99-102) claims that many loanwords in Table 2.3 were gaikokugo (4, p. 10), basing his argument on two factors which suggest a lack of morphological and syntacticallntegration (he also discusses a third factor, grammatical gender, but his evidence on this polnt is less compelling). Firstly, that some Catholic loans appear in both singular and plural form in surviving texts: e.g. both saNto 'salnt' and saJo;iosu 'saintS: As is well known, Japanese has no regular grammatical marker for the plural and gairaigo are typically shorn of any donor morphology (209, p. 141 ). Secondly, there are cases where a particle is lacking between an adjective and a following noun, e.g. supiricuaru koto ~,iritual matter: In the modern language, this would be supiricuaru no koto or supiricuaru na koto, and our knowledge of 16th century Japanese grammar argues for an expected, but unattested, ~supiricuaru naru koto.
13. The percentage of the Japanese population as of 2005 who identified themselves as Christian was appro.ximately 2% (STK 2008).
31
32
Loanwords in Japanese Table2.3. Selection of early Catholic gairaigo from Portuguese and Latin Japanese gaimigo (modern pronunciation)
English gloss
Source
First written attestation
Portuguese: trumaN batereN kiriSitaN
lay brother priest, father
aNjO
early Japanese Catholic angel
arutaru
altar
irmtio 'brother' padre cristao 'Christian
0 1568
•)1591
0 1569 ·:-1587
ewaNzeriyo
evar1gelist
iNheruno
hell
koNhisaN
confession
kurusu
cross
saNto(su)
saint
bapucizumo " baptttiz.umo paN
baptism
anjo altar evangelho infemo corifissito cruz santo baptismo
holy wafer > bread
pao'bread'
•)1591
missa sacramento gentio
0 1592
misa -- miisa
mass
sakarameNto
sacrament
zeNCiyo kiriSito (> kirisuto)
non·· Catholic, healhen Christ
pasukuwa
Easter
keredo
credo
maruciru
martyr
nataru
Christmas
osucia
host
rozario
rosary
histtpo
bishop
anima
soul
oraso
prayer
•)1591 •)1591 •:01591 •)1591 •)1591 •:01591 •)1591
•)1591 0 1593
'untamed, barbarian'
Cristo pascoa credo mtlrtir natal h6stia rosario bispo
01593 (>01886) 0 1599 0 1600 0 1600 0 1600 o-:-1600 o-:-1600 •)1636
Latin:
supir.icu
(holy) spirit
maNna
manna
katekizumo
catechism
deusu
god
abe maria
Hail Mary
ekerciia
church
hiidesu
truth
anima oratio spiritus manna catechismus deus ave maria ecclesia fides
•)1591 •)1591 •:01591 •)1592 0 1592 0 1598 o-:-1600 ·:-1600 0 1600
Chapter 2. A hlstory of Japanese loanwords
secret but no longer very well understood talismans of their religion: A few have clung to life, however. Ki1'isito, now kirisuto, is still the word for Christ, while rozario, misa, batereN and irumaN still survive, the last two familiar to many only from school history lessons. Even for modern-day practicing Japanese Catholics, however, many Iberian gairaigo have been supplanted by mostly Sino-Japanese words: batereN is now siNpu, koNjisaN is zaNge, bapucizumo is seNrei. Some of these words were originally Buddhist terms (zaNge, earlier saNge), while others are relatively recent neologisms (siNpu, 'lit. god father: is first attested in °1869; seNrei 'lit. washing rite' in °1870). Nataru 'Christmas' has been replaced by a loanword from English, kurisumasu [+1868]. In modern Japanese, a Christian is kurisucaN [0 1888], from English, or kirisutokyooto [0 1893 ], a hybrid compound. 'Catholic' is katoriQku [+1709] from Du. kathoiiek.
Ultimately, Mammon won over manna. It is, for the most part, not the Catholic loanwords in Table 2.3 which have survived, but those words for novel everyday items brought by missionaries and the merchants who followed in their wake. After the Ming opened Chinese ports in 1567, the Japanese did likewise and Portuguese merchants were allowed into Nagasaki in 1571. Later, the Spanish (1592), Dutch (1609) and English (1613) followed. Contemporary accounts paint a picture of a fascination, even an infatuation, on the part of the Japanese gentry and moneyed classes for the goods these merchants brought with them (Ishiwata 2001: 190---192). In time, however, the change to a political climate hostile to foreigners resulted in the English abandoning their trading post in 1623 and the Spanish being expelled the following year. The sakoku laws initiated in 1633 saw the Dutch and Chinese confined to Nagasaki by 1635, Portuguese traders expelled in 1639 and finally, in 1641, the Dutch, seen as unsullied by Catholicism and granted leave to remain, restricted even further to Dejima, an island in Nagasaki Bay. In §2.3, I examine in detail the huge contribution made to gairaigo history by these Dutch merchants, the only Europeans to remain in Japan. Meanwhile, Table 2.4 hints at the volume of secular loans borrowed from the Portuguese and Spanish before they were expelled. It is ultimately impossible to ascertain whether these words first entered Japanese through contact with missionaries or with mer-chants. Ishiwata (2001: 175) notes that until the late 16th century some daimyo did not deal with merchants directly but used missionaries as intermediaries. Most of the secular Iberian borrowings sampled in Table 2.4 tall into five broad semantic classes: f(lodstuffs (e.g. ciNta, kas-.ttera, teNpura); cloth and clothing (jubaN, kaQpa, ra5a); drugs and medicinal goods (miira, saboN, tabako); a small group of tools and appliances (jumsuko, kaNtera, karuka); and place names (igirisu,
33
34
Loanwords in Japanese Table 2.4. Selection of early secular gairaigo from Portuguese and Spanish Japanese gaimigo (modern pronunciation)
English gloss
Source
First written attestation
Portuguese: porucugaru > porutogaru karuta
Portugal
Portugal
traditional playing
carta
•}I 592 > *1625
card~
hiidoro
glass
tabako kaqpa
tobacco > cigarette
hiroodo sL-.,nyoro
velvet (chief) trader
gibao 'doublet' vidro tabaco cap a veludo senhor 'Mr., master, you'
miira
mynh > mummy
mirra 'myrrh'
kapitaN
captain
igirisu
England > UK
ra~a
woollen doth
capitao lngli!s 'English(man)' mxa (now feitro) 'felt' (vino) tin to (;Mo de) Castela
jubaN -- jibaN
undershirt
raincoat
CiNta
red wine
kasute(e)ra
type of sponge cake
0
0
1601 -- 0 1702 0 1603 0
1607
•}1608
()o)-1608 0 1609
1610 > 0 1631 •}1610 •}1613 0 1617 0 1625
0 1625
'(bread of) Castille' (now pao de 16) furasuko
flask, decanter
koNpeetoo
sweets, candy
marumero
quince
bot aN
button
waka kO(.)pU
beef
oraNda
Holland, Netherlands
jooro
watering can
suit of cups in cards
frasco confeito marmelo botao vaca'cow' copo Holtmda jarro 'jug, pitcher' calcador
karuka
ramrod (for gun)
teNpura
temperar 'to season; deep-· fried seafC.1od or tempero 'seasoning' vegetables (worthless) suit espada of spades > ugly woman Spanish:
subeta
sabON
soap
jabon
meriyasu
type ofknitwear
medias 'stockings'
kaNiera
hand-held oil lamp broom (plant)
candela 'candle' hiniesta
enisuda > enisida
•}1625 0 1625 0
1638
0
1651 •}1652
0
1655
0
1675
0 1680
0 1683 0 1748
0
1631 •}16//
0
1717
•}1719 > 0 1783
Chapter 2. A history of Japanese loanwords
oraNda, porutogaru). The bulk of them are still in use in the modern language. Some, such as botaN, tabako and teNpura, are of comparatively high frequency; a few, like waka and .~iNnyoro, have disappeared from the language. Still others possess, through later borrowing, a doublet \·\'hose usage is more frequent. Most of these doublets have been borrowed from English (e.g. reeNkooto 0 1912 'raincoat' for kaQpa, berubeQto +1871 'velvet' for biroodo), but occasionally from other languages (e.g. garasu 0 +1763, from Du. glas 'glasS, for biidoro). Use of the older Iberian loanword may now be restricted to elderly speakers (e.g. biroodo), have a limited semantic range (e.g. biidoro is generally confined to glassware terminology), or be restricted to dialect. The etymologies of some gairaigo in Table 2.4 remain problematic. I have listed saboN as a borrm\'ing from Sp. jab on (spelt xabon in the 16th century and pronounced \·vith initial [J]), but some scholars claim the source is Por. sabao 'soap: since a form saboN [+1708] is also attested . .Meriyasu is generally held to be from Spanish medias was and is typically pronounced me[o]ias and may have been perceived by Japanese speakers as me[ c] ias or similar - but some claim the source is an older Portuguese form meias. The source of kaNtera may have been Du. kandelaar or even Lat. candela, both 'candle'; in any case the word is attested late. KoQpu probably only has a Portuguese source in the sense of the suit of cups in cards (karuta) - the later sense of a drinking vessel is not attested until 0 1733 and is likely from Du. kop. There are alternative hypotheses as to the etymon of teNpura and the word is attested late: see the entry in Arakawa (1977). A few loanwords with a possible, though disputed, Portuguese or Spanish source are omitted from Table 2.4. 1hese have first written attestations postdating the Iberian period and include totaN [0 late 17th century] 'corrugated sheet iron', buraNko [+1698] '(children's) swing' and saboteN [0 1751] 'cactus'. Although, as noted, there \'\'ere also Chinese, English and Dutch merchants in Japan during the Iberian period, their contribution to gairaigo vocabulary appears to have been extremely limited.lJowever, Ishiwata (20()1: 187) states that the word gw·eborotaN 'Great Britain' is found in the lkoku Nikki, compiled between 1601 and 1655, and a small number of Dutch gairaigo (koohii 'coffee' ~1615, koQku 'cook' +1615,poNdo 'pound (weight)' 0 1637) are first attested between these same dates.
2.3
Dutch borrowings: Mid-17th to mid-19th century
The 220-year period between the expulsion of the Portuguese in 1639 and the opening of selected foreign ports in 1859 is characterised by rangaku .- f: or 'Dutch learning'. During this period, the Dutch were the only Europeans permitted
35
36
Loanwords in Japanese
to trade, or indeed have any contact at all, with Japan. 14 Although confined from 1641 to the island of Dejima in Nagasaki Bay, they were required to travel to Edo annually f()r an audience with the shogun. The only Japanese who had access to the Dutch merchants at their Dejima 'factory' were the local authorities and their interpreters, a position which was hereditary. The Dutch were forbidden to study Japanese. While the merchants' male servants from the Dutch East Indies were also accorded residence rights, no Dutch women \Vere allowed, although relations with local prostitutes were permitted. Japanese spies were ubiquitous. \l()s (1963: 342) asks why 'the Dutch put up with all these humiliations'. The answer, he notes, is that they made a tidy profit, at least initially. Latterly; after their balance of trade had begun to slide into the red, they enjoyed the prestige of being the only Westerners permitted to do business with the shogunate. Tangible trade was, from the Dutch side, mostly in silk, velvet, glassware, spices and sugar, much of it exported from the Dutch East Indies. The Japanese traded silver, copper, porcelain and lacquerware. The intangible trade in 'Dutch learning', as well as the knowledge gained about the outside world from the submission offosetsugaki, was also valued highly by the Japanese authorities. The periodic submission of these 'grapevine reports' was compulsory, although their content was frequently deliberately distorted by the Dutch (Jansen 1984:541---542). In the first eight decades or so of the period, however, any 'learning' was scant, since sakoku laws encompassed a ban on all \Vestern books. Knowledge could only be passed on via oral contact and thus dissemination was limited to Japanese interpreters and those who went on the annual journey to Edo. In 1650, for example, Caspar Schambergen, a Dutch East India Company doctor, was asked to stay on a few months and instruct the Japanese court physicians in medicine. In 1720, the shogun Tokugawa Yoshimune :reJ!)-t,~* (1684 1751) lifted the ban on secular western books. Japanese scholars were sent to Dejima to study the Dutch language, and the wholesale ingestion of Western sciences began. Paramount among these was medicine. In 1774, a group of rangaku scholars, headed by Sugita Genpaku {;1- l±J k a, published Kaitai Shin...,ho 'New Text on Anatomy: a
Despite being tolerated only under the strict supervision of certain designated feudal domains, trade with four other foreign entities was also sanctioned during this time: with late Ming and then Qing China, Korea, the Ainu and the Kingdom of Ryukyn. FUither, the 'Dutch' were not necessarily from the succession of states that equates to the modern Kingdom of the Netherlands. Philipp Franz von Siebold (1796-1866), mentioned helm¥, is a case in point. He was German, born in Wtirzburg, but had joined Dutch military service as a ship's doctor after graduating from university. Earlie1; the first Western scientist to describe the ginkgo tree, Engelbert Kaempfer (1651-1716), was also German, from Lippe.
14.
Chapter 2. A hlstory of Japanese loanwords
translation ofJohann Kulmus' 1734 Ontleedkundige Tajelen 'Anatomical ChartS, 15 despite most of them not knowing any Dutch at all when their work began. Indeed, the first detailed study of the Dutch language, Otsuki (1788), was only published 14 years later. 16 The first Dutch-Japanese dictionary, Inamura ([1796] 1997), was published after an earlier attempt at a dictionary ended when its compiler, Nishi Zenzaburt) W.f..:=. ~fJ. died in 1768. An official government translation bureau was only established in 1811. Otori (1964: 266-268) lists some 47 medical works translated from Dutch into Japanese up to and including 1866 whose originals are clearly identifiable. 'In addition to these translations, however, there is an even larger number whose originals are not known ... [and] the number of medical books translated into Japanese during the 90-year period after 1770 is imposing indeed' (opus cit.: 268-269). Other Western fields of study transmitted via rangaku included physics, chemistry, mathematics, astronomy, botany, geography, engineering and ballistics. Throughout the rangaku period the Japanese authorities remained acutely aware of the dangers even indirect foreign contact could unleash. The historical record contains many examples of books being destroyed, of arrests, forced suicides, imprisonment, internal exile and of other forms of government repression (Jansen 1984: 544···546). Philipp Franz von Siebold, the physician credited with introducing clinical medicine to Japan from 1823 and founder of the Narutaki Juku in Nagasaki, was expelled in 1829 on charges of treason and spying after having obtained maps of the island of Sakhalin (Stephan 1971: 38). In 1838 a rcmgaku school, 17 the Teki Juku, was founded in Osaka by Ogata Koan ~t;~· :1tlifr, which was to be attended by many future Meiji Era luminaries, including Fukuzawa Yukichi :fJ*'tt"t. Otori Keisuke *-~~11'- and Sano Tsunetami fi.!f 'it~. The cultural, academic and political importance of the field of rangaku thus cannot be overemphasized and is well summed up by Vos (1963:346): '[t]he rise ofJapan as a great power in the second half of the nineteenth century is, to a large extent, attributable to the swift adoption of Western techniques and sciences. For this process the foundations had been laid by the rangakusha [rangaku scholars]'. As he also notes, Dutch remained the official language for international
15.
Itself a translation of the original, Kulmus (1722).
t6. The earlier Maeno ([ 1785] 1994) merely 'teaches the form of Dutch characters, explains the pronunciation. lists the Dutch words and their Japanese translation. and explains the translation of short Dutch sentences' (Numata 1964: 245).
Not the first such school. Otsuki Gentaku, above-mentioned author of the first Japanese study of the Dutch la.llb'Uage, founded the Shirando rangaku school in Edo in 1786.
17.
37
38
Loanwords in Japanese
negotiations in Japan until 1870. Rangaku scholarship is traced in greater detail than space here allows by Otori (1964), Doi (1976) or Jansen (1984).
'Ihe 'intangible' and 'tangible' trade mentioned at the beginning of this section calls for a division of Dutch gaimigo into two groups: medical and scientific; and mercantile and other. Many mercantile borrowings penetrated the daily lives of the Japanese of the time. On the other hand, most medical and scientific loans \'\'ere either concepts only fully understood by a select tew in the upper echelons of contemporary Japanese society; or objects rarely if ever glimpsed and known only from the study of Dutch translations of Western scientific books. Ishi,Nata (2001: 230···242) provides an overview of loanword use in the literature of the period, revealing the extent to which Iberian gairaigo were still being employed and the frequency and usage of Dutch gairaigo at the time. Saito (1967) lists approximately 700 Dutch gairaigo, while Vos (1963) lists 322, of which he claims around 160 were still in use in the early 1960s. Greater penetration is found in dialects geographically proximate to Dejima. Koga (2000), in his dictionary of Nagasaki gairaigo, records 1,520 words of Dutch origin. Table 2.5 is a selection of Dutch medical and scientific loans. Kcn·era was colloquially known as korori, from kom1·i to .~inu 'suddenly drop down dead'. lvlorumoQto is used in modern Japanese both for the animal Cavia porcellus, known in English as a guinea pig, and for the human subject of an experiment (hence its inclusion with the medical gairaigo). Marmots, a total different ani·· mal family; are maamoQto in Japanese, a later borrowing from English and a word which does not carry any experimental euphemisms. In Dutch marmot and huiscavia 'guinea pig' are frequently used interchangeably. Until the mid19th century; poNsu 'Nas, according to Vos (1963: 368), a cold remedy made from bitter oranges (daidai) and sugar. Later, the medicinal aspect was lost and poNsu was use to designate a simple juice. Hnally, rice vinegar and eventually soy sauce, kelp (koNbu) and other ingredients were added to create the popular dipping sauce now generally called poNzu. TI1e mercantile gairaigo in Table 2.6 can be divided into f.rmr broad semantic categories: food and drink (e.g. koohii, hamu, biiru), materials (zuqku, garas.A, gomu), leisure (raQpa, orugooru, daNsu) and maritime terminology (masuto, doQku, deQki). There are also foreign currencies (doru, poNdo) and place names such as doicu (oraNda '.Holland' had already been borrowed from Portuguese). Some place names occur only in hybrid compounds and are not listed in Table 2.6. These include jagatara-imo 'Jakarta potato; whose truncated form jagaimo is the modern Japanese word for 'potato'. Dutch gairaigo exhibit the first signs of the
Chapter 2. A history of Japanese loanwords
highly productive modern phenomenon of mora-clipping (§3.4): neru for jitraNneru, sent for seru.fi, mohi for moruhine 'morphine' and ciNki for CiNkicuu1·u. Table 2.5. Selection of medical and scientific gairaigo from Dutch Japanese gaimigo (modern prommciation)
English gloss
Dutch source
poNdo kaNfuru erekiteriSiteeto > erekiteru > ereki. tanunomeetoru supoito reNZU CiNkiCUUrll > CiNki korera (korori) koNpasu mesu ONSU pONpU pONSU > pONZU kateeteru toN retoruto semeNiO kina ekisutorakuto > ekisu anikari arukooru gasu karaqto gipusu pesuto _jigitarlsu reeNpa > ri.Npa sooda karuki aateru > eeteru
pound (weight) camphor electrici.ty
pond kamfer elektriciteit
thermometer pipette lens tincture cholera (navigational) compass scalpel ounce pump cold remedy > punch catheter ton retort cement cinchona tree essence, extract alkali alcohol gas carat plaster cast plague, Black Death foxglove, digitalis lymph soda bleaching powder ether
thermometer spuit lens tinctuur cholera kompas mes OtiS
First written attestation 0 1637 0+>1713
()of,1765 > +1779 > +1798 ++1768 0 1770
+1785 0 1792 > 0 1829
o1793 er.Iss4) +1797 +1798 +1798
0+>1798 pomp pons 'punch' 0+>1799 > nia katheter +1802 ton +1808 0 1810 retort 0 1811 cement kina 'cinchona tree (bark)' +1816 •)1819 > ·>1822 extract alkali +1822 •!•1822 alcohol gas ·>1822 0 1826 karaat •!•1827 gips 'gypsum' 0 1829 pest digitalis +1832 lymfe ·~1832 > 0 1872 soda +1832 0+>1833 kalk 'lime' 0 1833 > +1862 aether > ether
(Continued)
39
40
Loanwords in Japanese Table2.5. Selection of medical and scientific gairaigo from Dutch (Continued) Japanese gairaigo (modern pronunciation)
English gloss
Dutch source
seemi. magunesiumu
chemistry magnesium
asubesuto
asbestos
kobaruto niokeru
cobalt nickel
riciumu s utorikinline
lithium strychnine
moruhine > mohi
morphine
repura puracina
leprosy platinum
karusiumu harogeN
calcium halogen
cifusu morumoqto
typhus guinea pig
chemie magnesium asbest kobalt nikkel lithium strychnine morfme lepra platina calcium halogeen typhus marmot 'marmot'
First written attestation 0 1833 0 1834 0 1837 0 1837 0 1837
G->1837 +1837 0 1837 > 0 1902 0 1842
.,.1851 +1862 +1867 0 1869
.,.1872
The Dutch were already trading in Japan three decades prior to the mngaku phase, with koohii, koQku and poNdo all attested prior to sakoku. Indeed, some gairaigo frequently accredited an Iberian source may in fact have had a very early Dutch one: e.g. kaNtera (Table 2.4). Vos (1963: 371-373) lists a total of 20 'Portuguese and Spanish loanwords reinforced by [later] borrowings from Dutch'. These include gairaigo \·vhere the Dutch and Portuguese (or Spanish) forms are extremely similar, either because they are cognate, because Dutch borrowed from Portuguese or Spanish, because Portuguese or Spanish bor·· rowed from Dutch, or because the donor languages borrowed from an identical third source. Conversely, proponents of what is kno\'\'n as the gengotagensetsu ,J [if:$ jt~, 'multiple etymology theory' in the Japanese literature (Umegaki 1963: 156; Ishiwata 2001: 155---156) hold that a later borrowing was reinfixced by an earlier one. Dutch and two later donor languages, German (§2.4.3) and English (§2.4.4), are all Germanic languages and lexical similarities are many. With borrowings occurring at the end of the rangaku period, it is occasionally dit1icult to ascertain which of these three languages is the source. There exists the further complication, as already noted above with von Siebold, that some of the 'Dutch' physicians practicing in Japan were actually German. C~fusu, for example, may be hom Ger. Typhus, ri:Npa from Ger. I..ymphe. DoQku, sukoopu and kaN are all attested late and may be borrowings from Eng. dock, scoop and can. The last example is not, as folk
Chapter 2. A history of Japanese loanwords
etymology would have it, Sino-Japanese. This is a misconception stemming from an identical reading kaN for the Chinese character ~,an older form of {6-, with which the gairaigo may be written (254, p. 168). Table2.6. Selection of mercantile and other gairaigo from Dutch Japanese gairaigo (modern pronunciation)
English gloss
Dutch source
koohli
coffee
ko..1ku raqpa
cook trumpet
safi.uaN
saffron
pisutoru biiru
pistol beer
doicu
Germany (pane of) glass
koJ!ie kok roc,per sa.ffraar! pistooi bier Duits 'German' glas vet 'fat' letter 'letter' komma ham kameleon orge.'10:organ) biik slroop gom hop porzd cacao lamp dans flanel mast rar1sel doek dek serge dok pek, pik 'pitch' kan haak schop
garasu heQto
beeflard
reqteru
label
koNma
comma
hamu
ham
kamereoN
chameleon
orugooru
musi.cbox
buri(Q)ki siroqpu
tin(plate)
gomu hoqpu
rubber, gum
syrup hop
poNdo kakao
£, pound sterling
cacao
raNpu
lamp
daNSU
dance
fura(N)neru > neru
tlannel
masuto
mast
raNdoseru > raNdo
backpack, satchel
ZUQkU
canvas, sackcloth
deqki
deck
seruji > seru doqku
serge
peNki
paint can, tin
dock
kaN hoQku
hook
sukoqpu
shovel, scoop
First written attestation •)1615 ' 10 1615
0 1684
•)1705 0.1'1713 ()-)1724 0 1725
()-)1763 0 1774 0 1783
•)1788 0 1798
()-)1799 ' 10 1803
0+18ll +1815 ' 10 1822 ()-)1822 ' 10 1822
()-)1826 •)1831 ' 10 1831
•)1832 > 0 1872 ·~1848
•)1850 > 0 1877 ·~1854
0 1857
•)1864 > o->1897 0 1867 0
1874
0 1886 0 1892 0
1905
41
42
Loanwords in Japanese
The vast majority of the gairaigo listed Tables 2.5 and 2.6 are still used in modern Japanese, although 1·epura is now politically incorrect and has been replaced by the hybrid compound haNseNbyoo 'Hansen's disease: after Gerhard Hansen, the Non\'egian doctor who identified the disease's causative agent. Others not listed have all but disappeared from the language (doNtaku 'holiday' 0~1871, from Du. zondag 'Sunday'; doroNkeN 'drunk' ~1863, from Du. dronken) or were sub·· sequently supplanted by later loans, overwhelmingly from English. Such replace· ments include suupu 'soup' [~1869] for soQPu [0 +1787], from Du. sop; and bataa 'butter' [~1870] for bootoru [~1695], from Du. bater. Occasionally, later English borrowings did not supplant Dutch gairaigo but created semantically similar dou· blets. In the modern language, f()r example, we find next to gomu 'gum, rubber' the English loan cuuiNgamu 'chewing gum' [~1916], next to garasu '(pane of) glass' \·\'e find gurasu 'glass, esp. tumbler' [0 1903] from the English, and alongside biiru 'beer' we find bi(y)agaadeN 'beer garden' [~1956]. Other Dutch mercantile gairaigo fell out of use when Sino .Japanese calques were coined in the second half of the 20th century (§2.4): seemi is now typically kagaku 'chemistry' [0 1862]. Yet other Dutch terms appear never to have attained gairaigo status but were immediatelycalqued into Sino·Japanese: e.g. fN E1t byooiN 'hospital' [0 1787], lit. 'sick house', from Du. ziekenhuis 'hospital'; iJ.-:rif bohaN 'birthmark' [0 1815], lit. 'mother spot', from Du. moedervlek 'birthmark';~. )!{ kaiguN 'navy' [0 1868], lit. 'sea army', from Du. zeemacht 'navy'. In 1853, the American Commodore Matthew Perry sailed his 'black ships' into Uraga harbour (modern Kanagawa Prefecture), forcing the signing of the Kanagawa Convention the following year. While this convention opened the two ports of Hakodate (Hokkaido) andShimoda (modern Shizuoka Prefecture) to United States ships for fuel, food, water and consular rights, it was not until the later Harris Treaty was signed between the two in 1858, and various 'unequal' treaties between Japan and other foreign powers negotiated shortly thereafter, that sakoku ended. In the ensuing five years, four further ports, Yokohama, Kobe, Nagasaki and Niigata, along with the cities of Edo and Osaka, were opened to f.r>reign trade. In 1858, Fukuzawa Yukichi, pupil of the Teki Juku rangaku school mentioned above and founder in Edo of \·vhat is now Keio University; discovered on a visit to the port of Yokohama that he was unable to read shop signage, \·Hitten mostly in English (Fukuzawa 1899). He had, since his youth, been studying the wrong fi)reign language.
2.4
Western borrowings: 19th century to present
After the Harris Treaty of 1858 and the Meiji Restoration of 1868, the fear of colo· nization propelled Japan rapidly towards modernization and parity with the West.
Chapter 2. A hlstory of Japanese loanwords
One of the most significant government undertakings was the lwakura Mission of 1871-73, in which participated over a hundred important figures in politics, academia and the bureaucrac.y, including the future four-time prime minister, Ito Hirobumi -tf",i-if .X. 'Ihe mission spent 21 months overseas observing Western institutions, travelling through the USA, the UK, France, Holland, Italy, Germany; Austro-Hung-c:lry, Russia and elsewhere (Kume et al. 2002). As Gordon (2003:73) notes, 'this experience powerfully motivated the ensuing shopping spree in the mall of Western institutions, from central banks and universities to post offices and police forces: This 'shopping spree' was a major factor in the flood of loanwords that was to follow a period in gairaigo history known in the Japanese tradition as the seiyo -~ # 'Western' period. In the latter half of the 19th century; the range of donor languages reflected loosely the itinerary of the Iwakura Mission and the languages of the states on which Meiji Era governments came to model their national institutions: the law (Germany), education (the US and France), the army and police force (France and Germany) and the navy (the UK). By the turn of the 20th century, however, the balance had begw1 to swing relentlessly towards English. But it is to English loans that I turn later (§2.4.4). Three other institutionally important foreign languages, Russian (§2.4.1), French (§2.4.2) and German (§2.4.3), had begun to be studied in Japan earlier than English, earlier even than the lwakura Mission. Other Western languages have contributed to the Japanese gaira.igo pool in the modern period, although their influence has been minor. Italian and Latin are wortl1y of note, but since these two languages have also loaned to the major donor languages just mentioned, borrowing may not have been direct. Italian loans encompass mainly musical and culinary terms: opera 'opera' [0 1881], sarami 'salami' [0 1901] ,forute 'forte' [+1924], supageQti 'spaghetti' [0 1931], piza ~ piQCa 'pizza' [0 1955], esupureQSO 'espresso' [+1955], ce1'0 'cello' [0 l955),pasuta 'pasta' [+1975]. Besides the small number of Catholic gairaigo dating back to the Iberian borrowing period (Table 2.3), Latin has also provided a small number of loans in the modern period: e.g. apuriori 'a priori' [0+1874], ego 'ego' [•1889]. As their patterns of adaptation attest (§3.2), many names have been borrowed directly from Czech, Farsi, Polish, Spanish, Swedish and other languages. 18
18. Although not a 'Western' language, recent borrowings from Arabic, absent from both NKD and Arakawa (1977), should not be overlooked. lNtifaada 'intifada' from lintifa:4al. fataha 'Fatah' from lfatahl, and jihtlado 'jihad' from ld:~iha:dl are all examples whose patterns of adaptation point to direct borrowing.
43
44
Loanwords in Japanese
The sudden shift in outlook towards the West after the Meiji Restoration went hand in hand with a collapse in the prestige of China ..For a few decades, however, new Western concepts often continued to be transmitted by means of the language attached to this formerly prestigious culture. Thus, the influx of Western borrowings in the late 19th century was complemented by a phase of Sino-Japanese loan translations or calques, known as kango yakugo in the Japanese tradition. Some·· times calquing operated in place of borrowing, at other times calques replaced already existing Western gaimigo. Concepts calqued rather than borrowed include tecugaku tff~ 'philosophy' [0 1869] and hokeN f;,f,~ 'insurance' [0 1879]. Examples of Sino Japanese calques ousting existing gairaigo include giNkoo ~Ki~r 'bank' [ 0 1871.] f.ix baNku [+1867] from Eng. bank, and kagaku 1{.~ 'chemistry' [0 1862] for seemi [0 1833] from Du. chemie. There are also examples of calques surviving alongside borrowings. 'Ihese include the calque sa5iNki 'If Jl-m [0 1897] and the loan k(y)amera [0 1897], both 'camera'; and .~iNkoNryokoo %1·~~At [0 1899] and hanemuuN [0 1878], both 'honey1noon'. For greater detail on the background to this calquing phase and for further examples, see Saito (1977), Coulmas (1989) or Kaiser (1991). 19 2-4.1
Russian borrowings
Although Japan's modern borders with it are all maritime, Russia (and later the USSR) is unique from the Japanese standpoint in having been the only European power with which it has shared a land border. 2 For a total of 55 years (1855-1875, 1905-1920 and 1925-1945) Japan and Russia/USSR shared, on a north-south basis, the island of Sakhalin (Karafuto, earlier Kitaezo or Okuezo, in Japanese). The t
°
The government-mandated Loanword Committee established in 2002 proposed mainly Sino-Japanese replacements for loanwords (Table 6.2), thus heralding a possible calquing renaissance. At the time of writing, however, it is unclear how permanent this revival will prove to be.
19.
20. Ignoring the very brief wartime occupations of such countries as Burma ( 1942-1945), which saw Japanese territory border British India.
Chapter 2. A hlstory of Japanese loanwords
Although Russian as a foreign language has never been accorded the highest prestige in Japan, historical geopolitics dictated that its study began at an earlier date than that of French, German or English. Daikokuya Kodayii :;k J..l!i. Jt:k.:k., a merchant shipwrecked on the Aleutians in 1782, travelled through Siberia to St. Petersburg and gained an audience with Empress Catherine II. One of his works, .Daikokuya ([1793] 2003), contains the first, albeit small, Russian--Japanese lexi-con. In 1804, a Russian emissary, Nikolai Petrovich Rezanov, arrived in Nagasaki to request trading relations. His credentials were written in French, which nobody in authority could understand, and his request rebuffed. In what has generally been interpreted by historians as an act of revenge, one of Rezanov's lieutenants, Nikolai Aleksandrovich Khvostov, unsuccessfully attacked the Japanese settlement of Kushunkotan in southern Sakhalin in 1806, doing so again the following year, raiding the Kurile island oflturup (Etorofu) as well (Stephan 1971:45-48). This led the shogunate to order the Nagasaki interpreters to study Russian in 1809. One of these interpreters was Baba Sadayoshi .~ ~ j( \B, who was ordered to Ezo (Hokkaido) to learn Russian from the imprisoned captain of a Russian warship, Vasilii Mikhailovich Golovnin. His studies resulted in the first translation of a Russian grammar, Baba (1813). For greater detail on this and other early Russian dictionaries and language materials, see Hida (1977). From the mid-17th century, both the Russians and the Japanese had begun to explore Sakhalin, the former from the north, the latter from an Ezo to the south whose colonization had begun as early as the 12th century (Siddle 1996: 29). The situation remained fluid until 1855 when the two countries agreed to share the island, without designating an official north-south border, under the Treaty of Shimoda (written in Dutch). In 1875, Japan ceded the southern half of the island to Russia in exchange for the Kurile Islands. Two years previous to this, in 1873, there appear to have been only around 1,000 ethnic Japanese colonists on the island (Stephan1971: 62). Clearly. although there must have been contact between the two languages up to this time, the small number of settlers meant its scale was small. Apart from a five-year period from 1920 to 1925 when Japan occupied the northern half of the island, Sakhalin was officially divided at the 50th parallel between 1905, when Japan regained control of the south after the Russo-Japanese War, and the 1945 Soviet invasion. However, 'only one road intersected the frontier... and it remained unused ... With brief interruptions ... [the] boundary was sealed' (opus cit.: 86). This lack of social interchange on Sakhalin is reflected in the selection of Russian gairaigo listed in Table 2.7, where loanwords have been split into four chronological divisions. The first group contains two very early loans, bararaika and seeuci, both attested as far back as 1807. The second group dates from between the M.eiji Restoration of 1868 and the Russian Revolution of 1917. With
45
46
Loanwords in Japanese Table 2.7. Selection of Russian gairaigo Japanese gaimigo
English gloss
Russian source
bararaika seeuc.i
balalaika walrus
6anana:tiKa balalajka CMBY'I siivuti 'sea lion'
cuNdora rosuke caa --- caaru --- caari
tundra Russian czar
tundra pyccKMif russki~j n;apb carJ
0
samowaaru uoqka "' wotoka uocuka toroika
samovar vodka
caMoBap samovar votka
1905" +-1911 N°1919 •}1909
BO~Ka
0
1910N +-1927 ~ +-1928
troika
1'pOt1:Ka tro}ka
·~19ll
bolshevik soviet intellectual
6oJibmeBMKJ.t bolsevii~d 1 coBeT soviet JmTeJIJIMreun;Jul intteWigiencija 6opnl bor.{f) Ta:iira tajga rre •IKa pi e(.i ka MKpa lkra 'fish rot! Ko.MJfHTepH lwm!inVem KaMrrauMJI kampar1jija
0 1921
N
boruseblki sobie(Q)to iNterigeNc(i)a > iNteri
First wrilten attestation
TYH;I~pa
•)1887 0 1904
·~1921 0
1921 > 0 1930
0 1924 borscht 0 1924 taiga 0 1925 kind of stove 0 1928 salmon roe 0 1930 Komintern 0 1930 ) ·~1930 fund-raising campaign toocika concrete bunker, go TO•IKa toc}ka 'dot, point' •}1934 position ···············-·························-..································-·········-·······························-···································-·······················-········-········-········-········-····· noruma quota, allotted tasks uopMa norma 'production 01949 quota' •)1953 homecoming (from ~oMot1: domoj 'homewards' damoi camps) Co} I 955 rrnpo)l(l(J1 piiroSkji piroslki piroshki •)1957 supuutoniku sputnik CIIYJ'HMK ~putn.itk 0 1966 Ko.M6.HnaT kombiinat koNbinaato industrial complex 0 1987 uepecrpoji:Ka perjestrojka peresutoroika perestroika gurasunosuci glasnost macHOCTh glasnostt n/a
borusic.i taiga pecika ikura komiNteruN kaNpania > kaNpa
the exception of rosuke, a now obsolete derogatory term for a Russian, it contains mostly cultural loans. From 1905, when the Japanese took over the southern half of Sakhalin, newly registered immigrants were given abandoned Russian homes (Stephan 1971: 87). Samovars and vodka (as well as peeika, attested a little later)
Chapter 2. A history of Japanese loanwords
are integral to Russian culture and may have been objects left behind.21 Although 'vodka' is generally uoQka or woqka in Modern Japanese, older pronunciations, such as wotoka or uocuka, are still encountered. At least three variants for 'czar' are found: caa, caaru and caari. The lack of a final/ru/ or iri! mora in the first variant suggests this may have been a borrowing from a non-rhotic accent of English (Tables 3.4, 3.5) rather than directly from Russian. The third group runs hom the Russian Revolution to the outbreak of World War II. These consist mostly of Marxist-Leninist terminology (iNteri, borusebiki, kaNpania, sobie(Q)to), but also some Russian cuisine (borusici, ikura). Ikum is nm'\' the second most frequent Russian gairaigo in the modern language (Table 2.2), although because of its very native Japanese phonotactics many speakers are unaware of its loanword status. It has undergone significant semantic narrowing. In Hokkaido and Sakhalin bountiful supplies of wild salmon meant that these fish supplied the bulk of edible roe. The final group consists of those Russian gairaigo attested after 1939. Noruma is generally cited as having entered the language when Japanese POW s held in the Siberian gulag began returning home, the word for which was the then common, but now obsolete, damoi. Peres.Atoroika and g:~Ara.sunosuci entered the Japanese Ian·· guage in the 1980s at approximately the same time as they did English, although the latter, despite being listed in several other major dictionaries, does not appear in NKD and is too late for Arakawa (1977). 2-4.2
French borrowings
The study of French in Japan had its impetus in two almost contemporaneous political events. The first of these was the arrival of Rezanov in Nagasaki in 1804 and the subsequent attacks by Khvostov on Kushunkotan and Iturup, as related in §2.4.1. Not only were Rezanov's credentials in French, but documents written in French were left behind after the attacks, whose content the shogunate was anxious to ascertain (Numata 1964: 251). The second incentive for the study of French was related to political authority in the home country of the Dutch merchants at Dejima, trading on behalf of the Dutch East India Company and the only Europeans permitted to trade with Japan (§2.3). In 1796, the ailing Dutch East India Company was
21. In addition, Japanese fishermen, mostly from Hokkaido, were permitted to reside in the extreme south of the island during its Rus:>ian 'prison camp' period between 1875 and 1905 (Stephan 1971:65-82). A limited number ofJapanese workers were also permitted to work in Japanese coal and oil concessions in Soviet nmthern Sakhalin between 1925 and 1945 (opus cit.: 131). Both these contacts may also have been a source of borrowing and more research is required here.
47
48
Loanwords in Japanese
nationalized by the Batavian Republic, a I"rench client state proclaimed the previous year as the successor to the Republic of the United Netherlands. In 1806, Napoleon Bonaparte created a French puppet regime, the Kingdom of Holland, in place of the Batavian Republic. The Kingdom was incorporated into the French Empire in 1810, although only for three years, Napoleon's defeat at the Battle of leipzig in October 1813 leading to the liberation of the Netherlands later the same year. Thus, for 17 years (1796···1813), the Dutch East India Company, and their merchants on Dejima, were subject to influence from Paris. From 1808, the combination of these events led the shogunate to order the Nagasaki interpreters to study French which, not being in contact with any native speakers of the language, they did by having themselves taught by Hendrik Doeff; head of the Dutch factory at Dejima from 1803 to 1818. The studies of three of these interpreters eventually yielded the Fumnsu /ihan (c. 1814-1817), a basic French vocabulary and grammar. The first French dictionary to be published in Japan, Murakami (1854), was quadrilingual Japanese-French-English-Dutch, \·vhile the first truly bilingual dictionaries were the later Murakami (1859) and Murakami (1864). By the 1850s the shogunate \·\'as using Dutch army officers to train Japanese troops and, as Dutch military terminology was replete with French borrowings, many of these began to filter into Japanese as the earliest French gai· raigo. Since the Meiji Restoration, the bulk of I"rench borrm·\'ings have come from those spheres of culture at which the French have traditionally excelled, the arts, cuisine and fashion, as well as the units employed in the metric system, which Japanese authorities phased in from 1885. Table 2.8 offers a selection of French gairaigo divided into fi)ur chronological groups by first \·Hitten attestation. The earliest group (1850-1880) shows hm·\' the comparatively early start to French studies in Japan led to a number of written attestations either slightly predating, or f.r>llowing swiftly on the heels of; the M.eiji Restoration of 1868. Many of these are metric units, employed at first by the army. Other early Frenchgairaigo are related to clothing (zuboN, saQpo) or the arts (koNkuuru, deQsaN). The second chronological group, attested between 1880 and the end of World War I, consists mostly of cultural (baree, eeikeQto) and culinary (m·aka·· ruto, omurecu) terms. The no\'\' ubiquitous fast food dish omuraisu 'omelette stuffed with (tomato ketchup flavoured) rice\ a loanword compound composed of the truncated fimn (§3.4) omu and raisu (from Eng. rice), is not recorded until 0 1931. The third group, attested during the inter·'Nar years (1918··1939), is a mixed bag. It includes an early example, saboru, of ru··verbalization (198···201, pp. 138--139), as well as culinary borrowings, such as gurataN and kurowaQsaN. The final and most recent group includes biniim., frequently cited as being borrowed from English, but surely borrowed hom French. Mirufiiyu and mesena are so recent as to be neither listed in Arakawa (1977), nor accorded attestations in NKD.
Chapter 2. A history of Japanese loanwords
Table2.8. Selection of French gaimigo Japanese gaimigo
English gloss
saqpo
chapeau metre champagne champagne kilogramrne kilogram me jupon 'slip' trousers litre litre centimetre centimetre millimetre millimetre competition, contest concours rough sketch dessln 'drawing, design,
meetoru SaNpaN kiroguramu > kiro zuboN riQtoru seNcimeetoru > seNci mirimeetoru > miri koN kuuru deQsaN
French source
hat
metre
First written attestation
*>1850 +1860 *>1860 •}1866 > *1881 0
1870 *1871 •}I 871 > 0 1935 •}1871 > 0 1888 *1873 *1873
outline' kirorneetoru > kiro
kilometre
guramu
gramme
kilometre gramme
kuudetaa
coup (detat)
coup d'etat
+1883
bifi.tteki
steak
b~fteck
0 1883
ornurecu
omelette
0 1885
atorie
studio, workshop
jaNru
genre
a.Nkooru
encore
omeiette atelier genre cognac ala carte etiquette ballet corzsomm{ .filet encore
kurowaqsaN
croissant
croissant
saboru oodoburu
skip, miss (work, etc.) sabotage 'work-to-rule, idleness' hors-d'oeuvre hors-dhouvre
sioku
chi.c
ruporutaaju
documentary, reportage gratin
konyaQku
cognac
arakamto
ala carte
ecikeoto
etiquette
baree
ballet
koNsome
consomme
hire
fillet
gurataN guraJ)lpuri aNkeeto biniim
grand prix
01877 > *1932 0
1877
0 1890 ()-}1890
0 1897 0
1908
0 1909 0 1909 0 1909 0 1917
*1918 0 1919 0 1925 0
1928
chic reportage
0 1930
gmtirz grand prix
0 1936
0 1930
0 1938
questionnaire, survey enquete soft plastic, vinyl virzyle
(Continued)
49
50
Loanwords in Japanese
Table 2.8. Selection of French gairaigo (Continued) Japanese gairaigo
English gloss
mayoneezu > mayo
mayonnaise green pepper, airhead piment holiday, vacation vaamces sommelier, wine sommelier
piimaN bakaNsu somurle mesena mimfiiyu
French source
mayonnaise
steward corporate support of mecenat 'patronage of the arts' the arts millefeuille millefeuiile
First written attestation 0
1954 > n/a
0 1961
·>1963 0
1975
n/a
n/a
With the exception of saopo, all the loanwords in Table 2.8 are still in use in the modern language. Not listed are a few which are falling out of use, including abeQku '(dating) couple' [0 1941] from Fr. avec 'witll, and deseeru 'dessert' [+1884] from Fr. dessert. Some of these have been replaced by English terms: e.g. abeQku is now typically kaQpuru [0 1940] from 'couple: while deseeru is now dezaato [ 0 +1896]. Others have English doublets: e.g. bakedoN from 'vacation' [0 1975] competing with bakaNSu; the anglicized smvpeN [+1871] competing with saNpaN. Many of the French source words in Table 2.8 have also been borrowed by English, usually with an identical spelling. In some cases it is possible to show beyond doubt that the gairaigo entered Japanese directly from French, f()r either cultural reasons (e.g. no English speaking country had adopted the metric system in the late 19th century) or phonological ones (e.g. the anglicized pronunciation of 'genre' would likely have yielded Japanese "joNru rather than jaNru). In other cases, however, the ultimate source remains controversial: e.g. konyaQku, omurecu, eeikeQto, etc. Loans such as resutoraN 'restaurant' (see (177) on p. 126), as well as nugaa 'nougat' [ 0 1904], kureepu 'crepe' [0 1933], raNjerii 'lingerie' [+1959] and neg-urije 'negligee' [ 0 1961] are other such examples not listed in Table 2.8. 2.4-.3
German borrowings
Of all the European languages examined in this chapter, only German has held a kudos even remotely comparable to English, and even then only until the US occupation. As noted in §2.3, a few 'Dutch' physicians who taught medicine in Japan in the first half of the 19th century were actually German, among them Philipp Franz von Siebold. Despite having been expelled from Japan in 1829, in Europe von Siebold published a number ofbooks on the country, mostly on its flora and fauna. He returned to Japan in 1859 with a Dutch trading company and, now viewed by the shogunate in a more positive light, was appointed in 1861 as an advisor
Chapter 2. A history of Japanese loanwords
to the official government translation bureau in Edo. Here, one of his students was Ichikawa Itsuki -rp- )l) ~ 1.f, regarded by Doi (1976: 55) as being 'the initiator of German language study in Japan'. 22 In 1860, a school of Western medicine had been established in Edo by a group ofJapanese physicians and in 1869 was elevated to a college, the Daigaku T!5ko, which eventually became part of Tokyo Imperial University. The new Meiji government decided at this point to adopt German medical practice, consigning rangaku medicine to history, the 'hierarchy of [the] German medical education system ... probably [being the] most appropriate [for] the Japanese who had long been accustomed to the rigid tradition[s] of feudalism' (opus cit.: 58---59). From the 1870s, first German--Japanese (Yamamoto 1872) and then Japanese-German (Saida et al. 1877) dictionaries began to be published. A number of German professors were soon invited to teach at Japanese medical schools and university medical faculties. This, combined with Germany being the chosen model for the Japanese legal system, meant that German was not only the language of medicine, but the second language of academia in general until 1945. Umegaki (1963: 180) notes that German loanwords were used as an argot amongst medical staff to conceal diagnoses and observations of a delicate nature from patients: toriQperu 'gonorrhoea' [+1953] from Ger. Tripper, hoodeN 'testicle' [~1954] from Ger. Roden, kooto 'faeces, stool' [n/a] .fi·om Ger. Kot are all examples. Such argot still exists and, even today; many elderly Japanese physicians possess an excellent working knowledge of German. Table 2.9 provides a selection of major German gairaigo, all of which are still part of the modern Japanese language. Not listed are German gairaigo which have lost currency. Notable amongst these are beQto '(hospital) bed' [+1872] from Ger. Bett, which was replaced early by beQdo [0 1887] from English. As noted in §2.3, some late medical borrowings from Dutch may actually have been German (e.g. c~fusu 'typhoid' from Typhus, riNpa 'lymph' from Lymphe). Chronologically by first written attestation, the gairaigo in Table 2.9 may be divided into three broad groupings. The oldest are attested between the Meiji Restoration of 1868 and the beginning of the .First World War. Many of these are medical terms representative of older technologies (e.g. gaaze, kapuseru), with more recent medical discoveries (reNtogeN, wakuciN) not being attested until the second group, whose written attestations date from 1914 until the outbreak of
22. As an interesting historical aside, Doi also records that von Siebold brought a number of linguistics books Vlith him, including Bopp's Vergleichende Grammatik. These were buried by the Japanese autho.tities and the official catalogue only reproduced in 1936. ~Had Japanese scholars made use of this literature [at the time], the later course oflinguistic development in Japan would have been different' (Doi 1976:56).
.51
52
Loanwords in Japanese
World War II. This second group also includes a significant number of mountaineering (zai1·u, hyuQte), skiing (.~aNce, ge1·eNde) and socio--political terms (ideorogii, dema). Also notable in this period was a number of now largely obsolete student slang terms \•\'hich are not listed: e.g. saN 'a beauty, belle' [0 1917] from Ger. schi)n; riibe 'lover' [0 1924] from Ger. Liebe. One which has survived into the present day, ho"~Never, is arubaito and its clipped f.ixm baito. This is first attested in +1914 with the simple meaning of 'work>, but by 0 1925 a sense of 'academic work, academic publications: and then by 0 1934 'part-time (student) work>, its predominant modern sense, is recorded (see also Umegaki 1963: 163). Another \·\'ord not listed has a disputed etymology. ZeQkeN, initially 'numbered saddle cloth (in horse racing): later 'a competitor's number tag in athletics' is first attested in °1930. This may be either from the verb decken 'to cover' or the noun Zeichen 'sign, identification', but adaptation is problematic in either case. The third group, those attested since 1939, are semantically mixed, although here too medical (kande, noirooze) and mountaineering terms (.~uraa..fr-tzaQku, aizeN) are found. Many of the most recent are industrial acids and not listed here. These are typically found in hybrid compounds, the loan suffixed with SinoJapanese saN 'acid': e.g. asupamgiNsaN 'aspartic acid' from Ger. Asparagin(siiure). Table2.9. Selection of German gairaigo Japanese gairaigo
English gloss
German source
kurorohorumu kookusu
chloroform coke (coal)
i.NpoteNCU > iNpO
impotency
Chloroform Koks lmpotenz Seminar Tuberkuiin Br1ergie Miirchen Gaze Kapsel .Allergie Tr:ema
zeminaaru > zemi
seminar
cuberukurL"'
tuberculin
enerugii
energy
meruheN
fairy tale
gaaze
dressing, gauze
kapuseru
capsule
arerugii
allergy
teem a
topic, theme
karuteru
cartel
arubaito > baito
part--time job
reNtogeN
x-ray
ryuqkusaqku > ryuqku
rucksack
wakuciN
vaccine
hyuqte
mountain(eering) hut
Kartel Arheit 'work' Riintgen Rucksack \lakzin H.Utte 'hut, cabin, lodge'
First written attestation 0 1877 0 1877 +1880> 0 1934 +1886 > •}1954 0 1893 0 1895 0 1899 0 1899 0>}1901 ·~1910
0 1910 0 1914 •!•1914 > •)1954 0 1918 01924 > 0 1935 0 1928 0 1930
(Continued)
Chapter 2. A history of Japanese loanwords
Table 2.9. (Continued) Japanese gaimigo
English gloss
sa.Nce
ski jump
German source
Schanze demagogy, false rumour Demagogie
dema
First written attestation 0 1930 0 1930
'demagoguery' ideorogii
ideology
kariesu
caries
teeze
thesis
zairu
climbing rope
ruNpeN
tramp, bum, vagrant
horumoN
hormone
gereNde
ski slope, piste
geru
(chemica]) gel
boNbe
cylindet~ cannister
Ideo Iogie Karies These Seil Lumpetl Horman Geliinde ....., lre• Bombe 'bomb'
0 1930
Dachshund Karte Virus Neurose Schnorchel Schlaftack Onanie Messe (Steig)eiserz Urethan Bisbahn 'ice rink' Gewalt
01942 0 1949
0 1930 0 1930 0 1930 0 1931 0 1934 0 1935 ·~1935
01935
·-··········-··········-··········-·········-··········-··········-·········-·········-·········-·········-·········-·········-·············-·············-·············-..····..································ daQkusufu:r.i.o
dachshund
karute
(doctor's) card
uirusu
virus
noiroo:l'.e
neurosis
sunookem
snorkel
sura(a)fuzaqku > sura(a)fu
sleeping bag
onanii
masturbation
InCQSe
trade fair (centre)
aizeN
crampon
uretaN
urethane
aisubaaN
frozen road surface
gebaruto > geba
violence, force
2-4-4
•}1950 0 1953 •}1953 ·>1954 > •1'1954 •}1954 •}1955 0
1956 •}1961 ·>1962 +1969> •}1969
English borrowings
As already mentioned in §2.2, the English opened a trading post for a decade between 1613 and 1623 at Hirado (modern Nagasaki Prefecture). Compared to the Portuguese, however, English merchants and the English language held minor status, and this only briefly. Only a very few loans are evident from this period, e.g. gureborotaN 'Great Britain' (lshiwata 2001: 187). As was the case with both Russian and French, it was fear of external threat that initiated the study of English in Japan, in this case the trigger being the Phaeton incident. One con· sequence of the creation in 1806 of the Napoleonic puppet state, the Kingdom of Holland, was that Britain was now at \'\'ar with the Dutch. In 1808, the British frigate HMS Phaeton, flying false Dutch colours, entered Nagasaki harbour,
53
54
Loanwords in Japanese
captured some Dutchmen and threatened Japanese shipping. Although she left two days later, the incident led the chief magistrate of Nagasaki to commit suicide and the shogunate to order the Nagasaki interpreters to study English the following year. Their studies generated the first English-Japanese dictionary, Motoki (1814). Nevertheless, the prestige still held by the Dutch language is evinced by the fact that the first Japanese --English dictionaf)', Hepburn (1867), did not appear until over half a century later. Meanwhile, the US had begun large-scale whaling in the North Pacific. In 1848, an American sailor, Ranald MacDonald, had himself put to sea in a small boat near the island of Rishiri off extreme northwest Hokkaido. He was captured, imprisoned and eventually sent to Nagasaki where he was forced to teach English to some 14 interpreters. One of these, Moriyama Einosuke ii!f- LLI %'-Z..!W, was ordered to Edo for the negotiations that followed Commodore Perry's arrival in Uraga in 1853 (§2.3). The opening up ofJapan which k1llowed saw an increased awareness of the importance of English. In 1873, a leading intellectual and later Minister for Education, Mori Arinori ~:ff :tL, seriously suggested adopting English and abolishing Japanese (Mori 1873). The previous year, the new Meiji government had promulgated the Education Law (gakusei), in which English was a compulsory subject to be taught in all middle and elementary schools. 23 It also set about recruiting nearly 2,300 foreign advisors (oyatoi gaikokujin) to aid with the adoption of \\/estern industry, science and technology, as well as the study of Western art, literature, history and philosophy. Of these, approximately 40% (928/2299) vvere British, 16% (374/2299) American and 1% (21/2299) Australian, meaning a clear majority (58%) were English speakers.24 Their number included such notables as the mining expert and amateur linguist Benjamin lyman (§4.3.2), as well as Japanologists Basil Hall Chamberlain and Lafcadio Hearn. By 1874, there were, according to government statistics (Umegaki 1963: 73), some 91 foreign language schools established in Japan, of which 82 were for the study of English. As in its relationship with both Portuguese and Dutch, the dialect of the port city of Nagasaki contained a larger number of English gairaigo. The same was
23. Initially, only attendance at elementary schools was compulsory. Full enrolment rates were not achieved until the 1920s (Benson & Matsumura 2001: 137).
24. The statistics cited are from UHABKS (1975:493) and are for oyatoi gaikokujin hired between 1868 and 1889. There is also a substantial number of foreigners listed under the 'nationality other: 'nationality unknown' and 'multiple nationality' categories, indicating that the figure of 58% English speakers is a conservative one. Other major nationalities were French (11 %), Chinese (II%) and German (8%).
Chapter 2. A history of Japanese loanwords
true of Kobe and Yokohama. Daniels (1948) and Inoue (2007) provide an analysis of the Japanese--based sailors' pidgin of Yokohama, which the former terms a 'port lingo' and the latter considers a 'restricted pidgin'. Both analyses are based on a brief tongue-in-cheek guide to the pidgin, Atkinson (1879). Daniels cites a number of English words which, doubtless due to Yi)kohama's proximity to the capital, spread to the standard language. These include hoteru 'hotel' [+1.850], hausu 'house > greenhouse, glasshouse' [+1864], koNsiroo 'consuf [0 +1871] and, from what he terms 'overseas English', booi 'bellboy; waiter' [+1873]. Park (1987: 30-31) and Kobayashi (2009: 23-25), in analyses of loanwords in a novel by Tsubouchi (1885), cite, among others, the following English gairaigo student slang terms, now all generally obsolete: emu 'money: oorudodeQto 'old debt', uiikuneQsu '(psychological) weakness: hii 'he: paasueeto 'persuade: saQparu 'supper'. Most of these are probably best defined as gaikokugo rather than gairaigo. Others cited by .Park and Kobayashi are still in use today, although they have typically undergone semantic change: goorudo 'gold: rnazaa 'mother: geQto 'get: puree 'play'. Japanese victory in the first Sino-Japanese War (1894-95) was tempered by the government having to concede to the diplomatic demands of Russia, Germany and France in the Triple Intervention of 1.895. Despite having gained the Liaodong Peninsula under the terms of the Treaty of Shimonoseki which ended the war, Japan was pressurized into returning it to China in exchange for financial reparations. Th.e public outcry which followed the Intervention caused a backlash against the three European powers concerned and, since neither the UK nor the US were directly involved, may be viewed as having strengthened the position of English as the major foreign language in Japan at the turn of the 20th century. At the same time, the US sport of baseball began to become popular in the archipelago and loanwords such as beesubooru 'baseball' [0 1889], auto 'out' [0 1896] and seefu 'safe' [01906] are first attested. Early English loans display patterns of adaptation somewhat different from those seen today. They may show an unexpected epenthetic vowel (§3.2.2.1): v ) v ( s.. h1rt (* sato v ) , tJa • k·ecu 'b uc ket, ('"/ . ) . ,see buras1 'brusI1, (.not *burasu., sacu , · Jak.eto Yamada (2005:84-93) and (135, 137; pp. 109-110) for further examples. Some are auditory loans (§3.2) and exhibit deletion: e.g. waisacu 'shirt: haNkaCi 'handkerchief: roosu 'roast meat', bisuteki 'beefsteak' (see §3.2.3 f()r attestation dates). Others are likely the product of contact 'Nith non--standard pronunciations: e.g. kyabecu 'cabbage' [0 1900], but earlier kyabeeji [0 1891 ]; rneNci 'mince(meat)' [ 0 1905 in the compound meNcibooru 'meatball'), but earlier miNsi [+1880] or miNsu [+1884]. While British English was in the ascendancy until the end of the 19th cen·· tury, growing global Americanization thereafter saw US varieties of English V·
0
,
55
56
Loanwords in Japanese
begin to dominate, despite the growing political tensions between the US and Japan which would eventually lead to war in 1941. Particularly pronounced in the first few decades of the 20th century >vere borrowings from the spheres of sport (riigu 'league' 0 1907,faN 'fan' ·>1920), music (metororwomu 'metronome' 0 1910, jazu 'jazz' 0 1924), politics (anaakL~uto 'anarchist' 0 1901, suroogaN 'slogan' 01924) and fashion (burausu 'blouse' 0 1925, suucu 'suit' 0 1935). For greater detail on English borrowings up to the mid··1920s, see Kabajima (1984) or Sago & Hida (1986). However, this influx ofborrowings was punctuated by government propaganda campaigns against the English language. These began with the Japanese invasion of China in 1931, which led to the outbreak of the Second Sino-Japanese War in 1937, when both the UK and the US began offering assistance to the Kuomintang resistance. After the Japanese attack on Pearl Harbor in December 1941, and the invasion by Japanese f()rces of British colonies and territories in south and southeast Asia the following year, the Americans, the British and British Commonwealth countries became enemy nations. English was branded a 'hostile language' (tekiseigo). Although no laws were officially promulgated outlawing its use, intense propaganda campaigns saw to it that English was widely, though never completely, boycotted.25 Roman letters disappeared from street signs, POST from post boxes and WC from public toilets (Umegaki 1963: 92). Magazine and company names. as well as cigarette brands (Satake 1981:47), were frequently de-anglicized. Alternative, mostly Sino··Japanese, 26 terms were proposed and quickly employed in place of English gairaigo. When Singapore was taken in 1942, its name was changed to soonaN 'Southern Shc3wa (soawa being the emperor Hirohito's reign name), while the nihoN arupusu 'Japanese Alps' were designated the cuub-u saNgaku 'central mountains' (Shillony 1981: 148). Some 'hostile language planning' examples are sho\\<'11 in (5). Such planning was particularly active in radio broadcasting (Sa-d) and in sports (Se-i), especially baseball. Sino-Japanese replacements were, occasionally, specifically coined loan translations, highlighted in bold. In the main, however, they were pre-existing, albeit sometimes relatively recent, words.
2.5. The only working language possible for the 194 3 Great East Asia Conference in Tokyo was English (Shillony 1981: 150). 26. The fact that Japan's also being at war with China seems to have been ignored by language planners in government propaganda departments may be seen as demonstrating how thoroughly the Sino-Japanese vocabulary stratum had been nativized. Alternatively, it may be viewed as a logical outcome of Greater East Asia Co-Prosperity Sphere propaganda which claimed that the Chinese were not being 'fought' or 'invaded: but rather 'saved from the West:
Chapter 2. A history of Japanese loanwords (5)
a. nyuusu b. anauNsaa c. rekoodo d. maiku(rohoN) e. saQkaa f. gorujil g. marasoN h. sutoraiku
auto paa.ma.
'news' 'annoWJcer 'record' (.tnie
'football' 'golf' 'marathon' '(baseball) strike' 'out' '.t->erm'
=> hoodoo :ffi.it => hoosooiN ffi: it Jil => on baN ;H£ ·.-.-.-> soowaki it ;lii· ~ => fuukyuu lft'\ljt => dak;mu tr ~. . . .> taikyuukyoosoo lll A~;t =>yo f. =>hike ·.-.-.-> tieNptiCII il,~
'intelligence, repolt' 'broadcaster' lit. 'sound disk' lit. 'transmission vessel' lit. 'kick ball' lit. 'hit ball' lit. 'endurance race' 'good, OK' 'lost' lit. 'electric hair'
Japan's surrender in 1945 was followed by the US Occupation, which lasted until1952 (1968 in the Bonin Islands, 1972 in Okinawa), the only time in recorded history Japan has been occupied by a foreign pmver. Despite the government attempts just described to undermine the position of English during and shortly before World War II, the status of English rose to even further heights thereafter. Some 5,500 US bureaucrats administered the country and initially around half a million US troops were stationed throughout Japan. English gairaigo dating from this period reflect the times: jiipu 'jeep' [0 1947], puraibasii 'privacy' [0 .1949], saNgurasu 'sunglasses' [•1949], sutoriopu.~oo 'strip sho'N' [0 •1950], disukujoQkii 'disc jockey' [•1954], jiiai 'GI' [0 1955]. The country was gripped by an English craze. Ogawa (1945), an English conversation textbook, sold an astounding 3.6 million copies in just four months at a time of extreme austerity. Pidgins such as 'bamboo English' appeared (Norman 1955; Miller 1967: 262···263). Since the end of the US Occupation, English loans have continued to Hood in unabated. Ignoring words for new technologies, the follm·ving are but a brief taste of this deluge: supoNsaa 'sponsor' [•1952], seku5ii 'sexy' [•1956], sutoresu 'stress' [•1957], iNsutaNto 'instant' [•1960], re}aa 'leisure, recreation' [0 •1961], noohau 'know··how' [•1962], niizu 'needs, demands' [•1975], ruucu 'roots' [ 0 1977], daieQto 'diet' [0 1981]. Many are so recent as to have neither an NKD nor an Arakawa (1977) attestation: hiiriNgu 'therapy' (from 'healing'), karuto 'cult: 1·ikw·uuto 'recruit, gaidomiN 'guideline, raijus.ttai1·u 'lifestyle: jaNku.fr-tudo 'junk food: misumaQci 'mismatcll, iNseNtibu 'incentive: akauNtabiritii 'accountability: Jasutofuudo 'fast food'. In an echo of the 'foreign advisors' phase of the late 19th century, the Japanese government inaugurated the Japan Exchange and Teaching (JET) Programme in 1987, bringing foreign teachers to Japan to work in and for the state school sys·· tern. By the late 1990s, the programme had a budget of US$400m (McConnell 2000) and in 2010 nearly 95% of the participants came from English-speaking countries (JET 2010). Since 2002, English has been a required subject in Japa·· nese junior high and high schools. Even bef(ne this date it was the de facto foreign language of choice (Gottlieb 2008: 144-145) and is currently compulsory for
57
58
Loanwords in Japanese
first-year students at the vast majority oflapanese universities. Under the primeministership of Obuchi Keizo i}· ).¥~ ;~ ~~ ( 1998--- 2000), the government even went as far as holding discussions on the possibility of designating English an official public language. In spite of this, Japan still ranks low in English international test scores (Gottlieb 2005:67 -69; ETS 2009). More than a quarter of a century ago, Stanlaw (1982: 171) wrote: 'few Japanese actually speak English well enough to converse with ioreigners beyond a rudimentary exchange of greetings (emphasis original). 1his still holds true today. One mark of the power of English loanwords is that they are not, as with other donor languages, overwhelmingly borrowed from nouns and verbs, but also from adjectives and adverbs (haQPii 'happy: raQkii 'lucky: riizonaburu 'reasonable: aNbariibaburu 'unbelievable'), interjections (ookee 'OK: noo 'no: iesu 'yes: baibai 'bye-bye: saNkyuu 'thank you: oomi 'alright'), prepositions (auto 'out: ofu 'off: dauN 'dowrl, aQpu 'up'), numbers (waN 'one: cuu 'two: surii 'three: ji:wsuto 'first'), pronouns (mai 'mY, ooru 'all'), prefixes (noN 'non-: iNtaa 'inter-'), and even articles (za 'the') and conjunctions (aNdooa 'and/or'). 1hey have permeated the language to such an extent that they constitute some of the first words children acquire (mama 'mummy; papa 'daddy') and are used as basic colour terms (buraQku 'blacK, guriiN 'green: buruu 'blue). They are beloved of bureaucrats and politicians for their powers of obfuscation and mendacity (naNseNsu 'nonsense: nookomeNto 'no comment: peNdiNgu 'pending'). The huge volume of English gairaigo f(>Und in the modern Japanese language necessitates setting aside the approach adopted elsewhere in this chapter of offering a 'selection'. Instead, Table 2.10 shows the 20 most frequent English gairaigo in magazines, extracted from NINJAL (2006a). The two most frequent, taipu and seQ to, both have frequencies of approximately 1 in 3000 words. NINJAL (2006a) provides separate frequency figures for 'magazine articles' and 'magazine advertisements. Advertisements in Japan are '[a]mong the domains of Japanese mass media ... where foreign elements are most extensively applied' (Haarmann 1989: 85). 27 If instances of gairaigo occurring in advertisements are filtered out, the overall ranking in Table 2.10 alters significantly, as shown by the number in paren -theses in the left-hand 'rank' column. Notable now is the fact that tere,faQkusu and biru plummet to 121st=, 136th= and 60=. These three words are used heavily in address information and 87%, 81% and 62%, respectively, of their occurrences are
27. English here is often employed for meta-linguistic purposes, promoting modernity or sophistication (Takashi l990b) and triggering ethnocultural stereotypes (Haarmann 1984). The general public may be unable to 'discriminate between the English promoted by advertisers and copywriters and the standard English taught as a subject in the school curriculum' (Gabbrielli 2005: 80).
Chapter 2. A hlstory of Japanese loanwords
in advertisements. SeQto, saizu and karaa, all found frequently in fashion advertisements, also drop down the ran.kings appreciably. Taking their place are ciimu 'team' (8th), purezeNto 'present' (15th=), kur~u 'class' (17th) and sutairu 'style' (19th). The majorityofloanwords in Table 2.10 are attested for the first time in writing between the Meiji Restoration of 1868 and the outbreak of World War I, although hoteru is i(lUnd as far back as 1850, probably a product of the Yokohama pidgin discussed above (Daniels 1948). Table 2.10. Twenty most frequent English gairaigo in magazines (extracted from
NINJAL 2006a) Rank
Japanese gairaigo"
Engli..<1h sou.rce
l1'equency
First written
(%o)
attestation
type
0.40969
0 1875
SeQ tO
set
0.40520
(terefoN >) tere
telephone
0.39959
4 (20)
saizu
size
0.34908
0 1922 (0 1878 >) n!a 0 1933
5 (2)
eNjiN
engine
0.32438
+1871
6 (5)
modem
model
0.28173
0+1894
7 (10=)
hoteru
hotel
0.27163
+1850
8 (32)
karaa
colour
0.26265
0 1914
9 (136=) 10 (9) 11 (3=)
faQkusu
fu.x
za- ji dezaL-..r
the design
0.26041 0.25929 0.25704
12 (13)
sisutemu
system
0.25480
13 (12)
supoocu
sport(s)
0.25031
14 (10=)
peeji
page
0.24918
+1872
15 (14)
(suupaamaakeQto >) suupaa
supermarket
0.24245
(01950 >) +1963
16 (18)
slrlizu
series
0.23347
+1893
17 (7)
(terebijoN >) terebi
television
0.22898
18 (3=)
reesu
race
0.22337
0 1892
19 (6)
poiNto
point
0.22112
20= (24)
seNtaa
centre
0.21663
+t872 0 1898d
20= (60=)
(birudiNgu >) biru
building
0.21663
+1926
1 (1)
talpu
2 (15=) 3 (121=)
1980sb 1970s'' 0 1912 +1885 0 1900
(01930 >) O+J95l
• NINJAL (2006a) lists the F and .8 segments of acron}'IIls such as lF andBl as independent entries. 1hese are ignored. See Table 5.5. b Too early for Arakawa ( 1977), no attestation dted in NKD. Tite machine first appeared in its modern form in this decade. c
Not listed in either NKD or Arakawa. The al:te!>tation is according to Ishino (1983: 159) and is for za, not in use as an affix (206, p. 1.40), but rather in fixed. slogans such as sufOQjlu za kcogai! 'stop pollution!'
it.~ main
d Tite 0 1898 attestation is for the baseball term, the main modern sense of'group of buildings where an activity is concentrated' not being recorded. until 0 1.949.
59
6o
Loanwords in Japanese
All the English gairaigo in Table 2.10 were borrowed from nouns or verbs, with the notable exception of za, whose allomorph ji is a recent development from the English emphatic the. The three most frequent gairaigo in NINJAL (2006a) borrowed from English adjectives are buraQku 'black' [+1870], orijinaru 'original' [+1883] and buruu 'blue' [0 1906]; from prepositions, iN 'in' [0 1906], oint 'of [n/a] and aQpu 'up' [0 1926]; and from pronouns, yuu 'you' [+1874], mai 'my' [+1953] and ai T [n/a]. The two most frequent hrbrid compounds containing an English gairaigo element are sukiijoo 'ski resort, ski area' [0 1935] and saradaabura 'salad oil' [0 1955].
Table2.ll. Twenty most frequent English gairaigo in the Mairzichi Shinbtm 1994-2003 (extracted ti·om Nakayama et al. 2007) Rank
Japanese gairaigo
English source
Frequency (meetoru
(terebijoN >) terebi
television
70.2
iraku ciimu (purosucicuuto, puroretariaato, puroguramu, purofesonaru puropagaNda, etc.>) puro (terorizumu >) tero
Iraq team prostitute, proletariat, (TV, radio) programme, professional, propaganda
69.3 68.4
6
guruupu
7
seNtaa
2 3 4
8 9 10
First written attestation
100) (0
1930 >) O.l-1951 0
n/a 1908
63.6
0 1886, 0 1926. 01930,.1932,.1934
terrorism
52.6
(01921 >) 01927
group
49.3
·>1882
centre
46.4
0 1898
Sl~utemu
system
40.5
·>1885
nyuuyooku pol.Nto
New York point
36.4 35.5
n/a +1872 0
11
saabisu
servi.ce, freebie
35.4
12 13
WaSL'
vVashington top
35.2 34.2
14
sutaato tai
start
32.'1
1930 n/a +1893 0 1902
tie
32.1
0
hotel
30.9
17
hoteru naNbaa
number
28.7
+1850 0 1899
18
meekaa
maker
27.4
0 1951
19
supoocu
sport(s)
27.2
0 1900
20
keesu
case
25.8
·:·1872
15 16
1908
Chapter 2. A history of Japanese loanwords
A different ranking is found if English gairaigo are extracted from Nakayama et al. (2007), an analysis of the digitized database of all the Mainichi Shinbun Tokyo editions published between 1994 and 2003 (c. 860,000 articles, 0.45 billion char·· acters). Their method of data manipulation does not, however, permit the citation of raw frequency rates, as in Table 2.10. Instead, I have normalized the number of occurrences of the most common loanword in their database (meetont 'metre' from Fr. metre) to a figure of 100, against which the number of occurrences of English gairaigo are measured. 28 Of these, the 20 most frequent are shown in Table 2.11. All the most frequent English gairaigo in the table were borrm\'ed from nouns or verbs, although three place names are listed. Three truncated forms (§3.4) also appear: ter· ebi from terebi}oN, tero hom terorizumu 'terrorism' and puro from various sources (attestation dates for puro are those of the truncated form). Only six \·\'ords appear in both Tables 2.10 and 2.11: terebi, seNtaa, sisutemu, poiNto, hoteru and supoocu.
2._5
East Asian borrowings
The prestige accorded Chinese culture throughout the long sweep of Japanese his·· tory has given way in the last century to a more neutral and at times even sinophobic attitude, with two Sino-Japanese Wars (1894-1895, 1937-1945) coming before the 1949 Chinese Revolution, after which diplomatic relations with the People's Repub·· lie of China were not renewed until1972. Although the island of Taiwan was part of the Japanese Empire from 1895 until1945, the people of the island 'came to be seen as an alien people who \-\'ere to be civilized under Japanese rule' (Vasishth 1997: 115). As of2005, nearly 520,000 citizens of the PRC, Taiwan, Hong Kong and Macao lived in Japan (STK 2008), approximately 0.4% of the total population. Research into Chinese gairaigo has been scant. This is due to their having been categorized by many as belonging to the Sino-Japanese vocabulary stratum, and to a tendency to equate gairaigo solely with Western loans. Before examining Chinese gairaigo, however, a brief excursion into the Sino·· Japanese stratum (§1.2.3) is nee·· essary. This stratum can be divided into three systems of kanji (Chinese character) 'readingS, corresponding to three waves of kanji borrowing - two main waves, the
Broadly speaking, Nakayama et al. (2007) divide their data into 'general' gairaigo, proper nouns and Roman alphabet acronyms. I have amalgamated the data for the first two of these groups but ignored the third gmup. Roman alphabet acronyms, which are not necessarily gairaigo, will be taken up separately in §5.4.2. Putthermore, a certain level of'entropy' (their term) had to be attained for a given gairaigo to be ranked by the authors. 28.
61
62
Loanwords in Japanese
kan-on J~i, followed by the later, minor Ui-in ~i-.29 By the completion of the go-on wave, derived from a Korean-peninsula--filtered 6th cen-
go-on~~- and the
tury CE topolect of the kingdom ofWu ~ (concentrated on modern Nanjing), the modern Japanese kanji corpus, ignoring later graphical amendments and simplifications, had been almost entirely imported. The second wave, the kan-on, officially adopted by the Japanese imperial court in the 8th century CE, was one of new readings for this corpus, broadly speaking adaptations of the Chang'an (modern Xi'an) topolect, then capital of Tang China. 'Ihe minor to-in wave is composed of a relatively small number of borrowed words and the new readings attached to the characters with which they were written. These were brought back from the Southern Song dynasty capital of Hangzhou by Japanese monks who had gone to the region to study Zen Buddhism around the 13th century. During the kan-on and to-in waves it was not new kanji which were borrowed, but rather new pronunciations f{1r those already in use, adapted from geographically and chronologically distinct Chinese topolects. This accounts for the fact that most kanji have more than one Sino-Japanese reading. 30 To return to Chinese gairaigo. A few have an orthographic source and these are read according to conventional go-on, kan--on or to--in readings: e.g. aheN P~ tf (or -r.t. h•.L-)'opmm · '(0 1713)f' •• (,or -~AI m'il h•.L. yaptan -- ·• )'opmm. · ' 31 Th· --ld ~'" . . rom_Be1.· lt;;rJJ--r"J n apwn _. ey sh. ou not be confused with the Sino-Japanese calques noted in §2A. Most Chinese gairaigo are auditory loans, though from which Chinese topolect they ultimately stem is sometimes problematic. \Vhile some may be written in native Japanese script, as are most other gairaigo, the majority are typically written in kanji. The readings for these kanji do not belong to any of the three Sino-Japanese systems just described - the go-on, kan-on and to-in - and this serves to mark them out as gairaigo. As an example, consider the Chinese loan gyooza ~f- 'steamed dumpling'. Here, the first kanji ~~ has a go-on reading kyoo and a kan-on reading koo. The second kanji -T has the reading si for both its go- and kan-on, as well as a less frequently found to-in reading su. The readings gyoo for tit and za for .::.r- deviate from the three kanji reading systems and thus gyooza f.k-f is conspicuously gairaigo. Although a kanji used to write a Chinese gairaigo may on occasion appear to be go-on, kan-on or to-in, this is merely felicitous. Consider meNcu W.::s- 'face, pride: an adaptation of Bei. mirlnzi and not attested until 0 19·13. The Beijing reading
29.
The ti'l-irt are also known as to-on~~~, so-on ;f._~, or so-in
*·=t.
30. For further detail on the history and development of kanji readings, see Numoto ( 1986) or Oshima (2006). 31. Since kauji readings are a dictionary tradition, such Chinese gairaigo are dictionary loans. See Figure 3.3.
Chapter 2. A hlstory of Japanese loanwords
micm of the first kanji W is adapted to meN, identical to its go-on reading meN (its kan-on is beN) . .But this is purely coincidental. No such coincidental adaptation occurs with the second kanji ::f· cu. Table 2.12 provides a selection of auditory Chinese loans. Semantically, the majority are related to Chinese cuisine or to mahjong. Although a kanji spelling is listed f(lr each, many may be written in other scripts (255, p. 1.68). Not included are hybrid compounds where one element is Sino-Japanese, such as aNzeNpai :tc-1:-W 'safe pair of hands' (lit. 'safe mahjong tile') or poNkaN ~;Itt 'Poona orange, citrus reticulata' (the elements in bold are borrowed). Hybrid compounds where one element is analysable as either Sino-Japanese or an auditory loan are included, however: e.g. uuroNca and maaboodoofu (elements in bold are borrowed, remaining elements may have been borrowed). 32 Although most of the gairaigo in Table 2.12 are noted as having been borrowed from the Beijing topolect, with a few from Cantonese, the possibility remains that a different topolect was the ultimate auditory source further research is required in this area. 'Ihe loanword gyooza 'steamed dumpling' is a particularly problematic case in this respect. Bei. jiaozi is not a particularly close phonological fit and alternative donor hypotheses include Hakka; a topolect spoken in Shandong .Province; and non .. Chinese sources, such as Manchu. Two gairaigo not listed in Table 2.12 are generally held to be borrowings from Chinese, but have problematic etymologies: peteN 'swindle, fraud' and caNpoN. The first is attested from 0 1874, but its donor word and topolect are unclear. The second is attested from 0 1761 with the meaning 'musical ensemble: from 0 1825 as 'mixture: and from 0 1949 as a fried noodle dish popular in southern Japan and Okinawa. Again, the donor word and topolect are unclear. Chinese names are, for the most part, orthographic borrowings. The kanji with which they are written (Figure 5.2) are read with conventional kan-on, go-on or to-in readings. This practice is known as wameihyoki ~4i ::ttc. and examples include koosuu 'GuangzhoU. kooga 'Yellow River: siseN 'Sichuan' and cuugoku 'China'. Since most Chinese place names have been known to the Japanese since before the mid-16th century, under my definition in (3) on p. 10 they do not belong to the gairaigo stratum but to the Sino-Japanese. Nevertheless, there do exist a number of Chinese place names, generally large cities. which are auditory loans and whose Chinese pronunciation has been adapted. Titis practice is known
In other words, the ca and doofu elements here may be the Sino-Japanese word'l ca
32.
63
64 Loanwords in Japanese
as hatsuontenki 16~· ~We. Examples include saNhai 'Shanghai: from Bei. shanghai L 'ifr • 'I·Iong Kong,) from cJant. neung . 1gon;s,,2 ir"l~. _,;. '.llo. _t..:mf, an d 1wNkoN Table2.12. Selection of Chinese gairaigo Japanese
Kanji
gairaigo
spelling
ii kooryaN uuroNca Wa:NtaN maajaN
one type of millet ~~.~ i.~ ~it#; oolongtea ~;:g: il.'!.i't wonton dumplings mahjong .iff.~
English gloss
Chinese source
Bei.yi~
Bei. gaolidng jli; 1*. Bei. wU16r1gdul Jt ;t;ljl; Cant. wan4 tar1 1 ~;if: Bei. mtijiang JW.~It maque
Nrst written attestation 0 1809 0 1907 0 1914 0 1914 0 1926
JW.~
pai suumai raameN
}~ tJ't~
tft~ j~-
BeL pai ,')¥ mahjong tile steamed dumpling Cant. siu1 maai6 ~~ Bei. lamiiln 'pulled noodle' noocUe soup
t.v.li caasuu toimeN caahaN meNCU raicti gyoo7.a jaajaNmeN ~ jaajaameN maaboodoofu piNiN puaaruca SOOfONpOO
toobaNjaN yamuca
J:.:KJf. ;0'a.'W.
.trW t.l>·~\l.
mi-T ,#.;.rt.
#.-f· :~:til 'AI
!tT~JL~
#tfftf-~
;J,iUz. :a~41-
t~~
roast pork, char sui Cant. cha 1siu 1 YJ!f, Bei. dltimitln 'person player opposite opposite' .tt 11ii Bei. chiiofan k}{IZ fried rice face, pride Bei. miiinzi W-'7 lychee Cant. lai 6ji 1 .
Bei. mdp6doufU M-~ :~LI~ Bei. pinyin~'* Bei. puercha 'IH·t-~ Bei. xiiiol6ngbao ;J,;lH!. Bei. douban}iiing li;!ltif Cant. yam 2cha 4 tx~
0 1926 0 1928 0 1930 0 1931 0 1932 0 1939 0
1943
0 1955 0 1955
n/a n/a n/a n/a n/a n/a n/a
Recent decades have witnessed a shift away from Sino-Japanese wameihyoki towards auditory hatsuontenki loans. Some of the better--known Chinese cities now have both a long--established wameihyoki name (listed first) as well as a newer, often less established, hatsuontenki one (listed second). Examples include taihoku ~ taipee 'Taipei' (Bei. taibei if ~L) and juukei ~- coNciN 'Chongqing' (Bei. ch6ngqing j[f;\). Place names are notoriously fickle, however, and many Chinese place names have had diiferent Japanese pronunciations in the past, while during the many centuries of Sino-Japanese contact cities have been completely renamed by the Chinese. Virtually all ancient Chinese personal names, as well as many recent ones, are wameihyoki and cannot be considered gairaigo: e.g. moo takutoo -"E;~F(-*. 'Mao
Chapter 2. A history of Japanese loanwords
Zedong: too soohei !t~ /j\ -'1'- 'Deng Xiaoping'. Nevertheless, as with place names, recent years have seen an increase in Chinese personal names with auditory hatsuontenki pronunciations, although most still have conventional wameihyoki doublets. This is illustrated in (6). The official guidelines of NHK, Japan's only public broadcaster, state that Chinese personal names should 'as a rule' be read as Sino--Japanese wameihyi:!ki (NHK 2006:40). (6)
hatsuontenki Japanese warnei11yi'lki spelling (Sino-Japanese) (gairaigo)
PRC Beijing spelling pinyin -~~1-~w ~~:;J·tt
IHt!fntao ~~~~ Gao Xfngjiim !ti4Ht
ko kiNtoo koo kookeN
fujiNtao gaosiNjeN
\Vith Korean personal names, auditory hatsuontenki loans are more common and began earlier, driven by two major factors. Firstly, the declining use of Chinese characters in South Korea, and their prohibition in the North (Lee & Ramsey 2000: 56-60), has meant Koreans themselves frequently no longer use them to write their names. Secondly, pronouncements on the part of the Japanese establishment advocating hatsuontenki. These have included NHK who, while being sued, ultimately unsuccessfully, by a Korean resident of Japan named-#!. -~:jf in the Japanese Supreme Court (SSS 1988), in the meantime acceded to the litigant's demand to have his name pronounced 'in the Korean fashiorl, i.e. adapted hatsuontenki, as coe caNhoa. The Japanese Foreign Ministry also received a similar request for hatsuontenki from the South Korean government prior to a presidential visit in 1984. The travel advisory section for South Korea posted on the .Ministry's homep·· age states that 'it is desirable that Korean pronunciation is used as far as possible for South Korean personal and place names' (Gaimusho 2009). No such exhortations are posted in the travel advisories for North Korea, the People's .Republic of China or Taiwan. NHK official guidelines (NHK 2006:40) state that the names of South and North Korean citizens should 'as a rule' be read as hatsuontenki, 33 while the same applies to Korean residents of Japan, unless they have expressed a wish otherwise. Some examples are shown in (7). (7)
Korean and Japanese spelling
1tan'gUJ
*-l~-t~J.
o] r.a1':J.r
t.f£jz: ~~.IE E
~t7l
¥.-
7J ::>.J ~I
Korean (Mc.CuneReischauer)
Yi Myong-bak PanKi-mun Kim ChOng-il
Japanese
hatsuontenki (gairaigo) i myoNbaku paNgimuN kimujoNiru
33· See l (1988) for general problems related to hatsuonteuki adaptation and its Japanese transcription.
65
66
Loanwords in Japanese
With older Korean personal names, \·\'e find wameihyoki/hatS"uontenki doublets as far as back as the introduction of the han'gul writing system in 1443 or 1444. This is illustrated in (8). In contrast to Chinese place names, Korean place names are now typically auditory hatS"uontenki, although Sino-Japanese wameihyoki pronunciations, whose usage is declining, also exist. Just as with Korean personal names, NHK broadcasting guidelines (NHK 2006:40) advocate the use of hatsuontenki with place names also. Some examples are shown in (9). Korean and han'gitl Korean (McCuneJapanese Reischauer) spelling (8)
(9)
j..!..:...;;;r.:; *~
;.~Hi"
-~J~.:MJ
;8~·'iJ.
-~w
11.~
1:::.)11
_;F:!l
~oJ=
l=l
.A]-
Japanese
Japanese
hatsuontenki (Sino-Japanese) (gairaigo) (King) Sejong seisoo sejoN kimuoQkyuN Kim Ok-kyun kiN gyokukiN
Pusan Jnch'On P'yc)ngyang
wameihycJki
fuzaN _jiNSCN he~joo
pusaN iNCON pycmyaN
If one ignores names, loanwords from Korean are comparatively few. One of the primary motivations for borrowing is the notion of prestige and Korean, despite its geographic proximity, has clearly sutlered. The Korean peninsula was a Japanese colony from 1910 to 1945, suffering acute linguistic imperialism- the Korean language had been banned in schools by the end of the period. Official Japanese relations with Korea across the bulk of the modern period have ranged from nil (during the period of the occupation) to very poor (its immediate aftermath, and also on a continual basis with North Korea) to poor. The first decade of the 21st century, however, saw much improved relations and a surge of interest in South Korean culture, with its popular music and television soap operas now commanding considerable media attention. Just under 600,000 citizens of North and South Korea lived in Japan in 2005 (STK 2008), a figure that has fallen from approximately 683,000 in Osaka, Hiroshima, I"ukuoka and Nagasaki alone in 1943 (Weiner 1997b: 87). As can be seen from 'Htble 2.13, many of the most recent borrowings from Korean are restricted to cuisine and few have attestations cited in either NKD or Arakawa (1977). Ainu is another example of a geographically proximate language which has suffered from lack of prestige. Despite continuous contact on Hokkaido between ethnic Japanese settlers and the indigenous Ainu, Ainu borrowings are extremely scant. Examples include tonakai 'reindeer' (~1804, fi-om tonakkay), kotaN 'Ainu village' (0 1936, from kotan), sisamo 'species of smelt' (0 1957, from S'usam), ruipe 'frozen salmon or cod sliced and eaten raw' (n/a, from ruipe) and iyomaNte 'brown bear sacrifice ceremony' (n/a, from iomante). Th.e legacy of linguistic contact in I::lokkaid6 thus remains overwhelmingly with frequently encountered Ainu place
Chapter 2. A history of Japanese loanwords
names such as Sapporo, Monbetsu, Wakkanai and Obihiro, as well as with the word ainu itself (from aynu 'human').
Table2.13. Selection of Korean gairaigo Japanese gairaigo
English gloss
Korean source
paoci kiisaN ~ kiiseN coNgaa cige cima
tight·· fitting trousers female entertainer bachelor spicy stew traditional Korean garb Korean letters traditional Korean garb pickled cabbage traditional Korean garb raw minced beef with egg won (currency) genre of popular music rice dish juche (officially state ideolo!,ry of North Korea) thinly sli.ced rib meat pancake, jeon barbecued beef dish taekwondo Korean alphabet, language
pajitl}7~!
ONlllON
cogor.i kimuci cimacogori. yuqke WON
poNcaqku hibiNba cuee kanibi ciJimi purukogi tekoNdoo haNguru
2.6
kisaeng 7] "'~ chbnggak ·%-4 tchigae ;J0<]7H ch'ima. 7.1 6mmmz!~·
ch6gori ?.12 i'-] kimch'i {] ,-• ch'ima)6gori ~1o} ~12 c1 yukhoe-,\f-~1
won-%!. ppongtchak :wozJpibimpap l:ll ~ ~ ch ucht: 'T· .:<1] kaibi ~\:ll chijimfl{] pulgogi *271 taekwondo
First written attestation 0
1722 1885 0 1895 +1910 +1928 0 1937 0 1946 +1948 0 1975 n/a nia n/a n/a n/a 0
nia n/a n/a nia n/a
Names
Non-kanji-sphere place names are typically orthographic borrm·\'ings. Spelling loans (Figure 3.3) are not uncommon: jamaika 'Jamaica: mNdoN 'London', suweedeN 'Sweden' and kuroacia 'Croatia' are but a few examples, all with an English source. Place names are, in the main, adapted from the language of what is, or was, the colonial power; or from what is, or has been, an official language of the state in which the place is situated. This does not ah\'ays correspond to the mother tongue of the majority of local inhabitants. Thus, pari 'Paris' from Fr. Paris, napori 'Naples' from It. Napoli, rnyuNheN 'Munich' from Ger. }vfiinchen, warusawa 'Warsaw' from Pol. Warszawa, sebiriya 'Seville' from
67
68
Loanwords in Japanese
Sp. Sevilla; but moNtoriooru 'Montreal' and kairo 'Cairo' from English, gurujia 'G_jeorgta . ' trom ~ Ru. gruz-l}a _;.. rl.p)TSHH,-' 4 kaza rfiusu t aN 'K aza-kh stan' f rom Ru. kaza-, xstan Ka3axcran, and kirugisu 'Kyrgyzstan' from Ru. /Jirgjis K11pr113 'Kyrgyz'. There are a significant number of place names where the donor language is English. Surprisingly, these include places where the local or state languages have been significant donors to Japanese: e.g. risuboN 'Lisbon', but Por. Lisboa; aNtowaapu 'Antwerp', but Du. Antwerpen and Fr. Anvers. Although a few other Englishdonor place names can be found within Europe (e.g. kopeNhaageN 'Copenhagen', but Dan. K0benhavn), they are more prevalent outside (e.g. baNkoku 'Bangkok: but Thai Krung Thep Maha Nakhon). English--donor place names are especially prevalent with countries: supeeN 'Spain', but Sp. Bspana;jiNraNdo 'Finland: but Fin. Suomi; arumenia 'Armenia', but Arm. Hayastan. The recent trend for a city once part of the British Empire to officially alter its name from the English to the local one has also been reflected in altered Japanese naming practices. What was boNbee 'Bombay' may now be muNbai 'Mumbai: while what was raNguuN 'Rangoon' may now be yaNgoN 'Yangon: There are a few place names where the donor language is neither a local language, an official language, a colonial language, nor English. These are mostly country names borrowed into Japanese very early: girisa 'Greece: oraNda 'Holland: igirisu 'UK: fi·om Por. Grecia, Holanda and ingles 'English(man)'; as well as doicu 'GermanY, from Du. Duits 'German'. Diachronically, foreign place names in Japanese, as in other languages, have been characterized by chronic instability. Many of those mentioned in the preceding paragraphs have had earlier forms no longer in use, while some still have several competing modern forms. In some cases, an earlier form has been fossilized in a compound but replaced elsewhere: e.g. zenebahaguruma 'Geneva gear: but _juneebu ~ juneevu 'Geneva, from Fr. Geneve. In others, a competing, often Englishderived f(xm, exists only in compounds: e.g. mosukmnyuuru 'Moscow mule: but mosukuwa 'Moscow: from Ru. moskva MocKBa. I close this chapter with a very brief review of gairaigo borrowed from brand and/or corporate names. Restricting discussion to those foreign brand and corporate names which have been genericized in Japanese (Clankie 1999: 184---216), examples include baNdoeedo 'Band Aid plaster'; kurakusoN 'car horn' from Klaxon; barikaN 'hair clippers' from the French company Bariguand & Marre; hoQcikisu 'stapler' from the now defunct American firm E. H. Hotchkiss Co.;
34· Following the conflict with Russia in August 2008, the Georgian foreign minister Grigol Vashad1..e made a formal request to the Japanese foreign minister Nakasone Hirolitmi tt' ~'* g _9:. in March 2009 that the Japanese name for his country be changed to the English-derived joojia (7he Times, 26 March 2009).
Chapter 2. A history of Japanese loanwords
doomN 'greasepaint' from the German company Dohran; and baikiNgu 'buffet, smorgasbord' from the Viking restaurant in the Imperial Hotel, Tokyo. Also worth noting are semantically remodelled (§4.4) English b.mnd names, devised and registered in Japan but composed of English elements. Genericized examples include saapupeNsiru or saapeN 'propelling pencil; both clippings of ebaaredisaapupeN.~iru 'Ever Ready Sharp Pencil; from which the Sharp Corp. took its name; pokeberu 'pager, beeper: a clipped form of Eng. pock(et)+bell and trade·· mark of NTT Communications Corp.; ·wosureQto 'super bidet/toilel, a clipped form of Eng. wash+(toi)let and trademark of the toilet manufacturer Toto Ltd.; and wookumaN 'walkmml, a trademark of Sony Corp.
69
CHAPTER 3
Phonology The phonology of]apanese loanwords is a heavily researched field. Although many recent studies have been theoretical in f(Kus, 35 in this chapter I aim to provide a comprehensive, but purely descriptive, analysis of gairaigo phonology. After a preliminary overview of issues related to phonemicization in §3.1, I move on to the core of the chapter in §3.2: issues of adaptation. Th.ese I divide into substitution (§3.2.1), epenthesis (§3.2.2) and deletion (§3.2.3). Suprasegmental issues are exam·· ined in §3.3 and the chapter concludes with an analysis of mora-clipping in §3.4.
:P
Phonemicization
Until approximately half a century ago, the typical speaker of standard Japanese displayed a conservative, or traditional, phonemic system. Such a system is still found with many elderly standard speakers today. An analysis of such a phonemic system - a conservative phonemic analysis - generally posits 16 consonants and the five vowels /a e i o u/. Although there is some divergence in opinion, a consonant phonemicization such as that in Figure 3.1 is broadly representative. 36 Under such analyses, the following allophones are typically posited: (10) /h/ is [(p] befixe /u/ and [<;: 1before /i y/; ( 11) It dl are [ts z1 before lui and [cs;. J'~ 1before Iii; (12) Is zl are [~> J?>] before /i/. In addition, appearing before /a o u/ only, /ty dy sy zy/ are [cq?> q?>], as shown in (13).
35· For example, Dupoux et al. (I 999), Shinohara (2000, 2004), Labrune (2002), Kubozono (2002a, 2006), Dohlus (2005b), Peperkamp et al. (2008), Mutsukawa (2009). 36. Analyses identical or similar to Figure 3.1 have been put forward by, among others, Rkkmeyer (1989:28-30), Rothaug (1991:73-80), Yamaguchi (2007:16) and 'Isujinmra (2007:22-37).
72
Loanwords in Japanese
plosive fricative nasal flap approximant
biklbial p b
alveolar t d s z
m
pakltal
velar k g
glottal h
n r
moraic Q N
y
w
Figure 3.1. The Japanese consonants: a typical analysis of a conservative phonemic system
(10)
lhuhool lhihool lhyoohoNI
[cpwho:] ['riho:] ['rjo:h6N:]
(11)
ltetul lhitodumal Itotil lhanadi/
[tetsw] [<;:itozwma]
(12) (13)
lsusi/ lzizeNI
ltyal lkutidyal lsyul lzyol
'illegality' 'sad tidings' 'specimen' 'iron' 'married woman' 'land' 'nosebleed'
[toc~i]
[hanaJ;i] [sw~i]
[J;izeN:]
'sushi' 'charity'
[~a] [kwc~iJ;a] [~w]
[J;o]
'tea' 'fresh tea leaves' 'type, kind' 'preface, foreword'
The influx of gairaigo over the past half century has had considerable impact on the conservative phonemic system, severely denting the authority of conservative analyses. The bilabial [
37· Similar analyses have been put forward by Vance (1987, 2008a), Labrune (2006:731f) and Tranter (forthcoming). For phonemicizations differing substantially from both Figures 3.1 and 3.2, including those which posit separate analyses for individual vocabulary strata (i.e. coexisting phonemic systems), see Bloch (1950), Martin (1952: 12-15), Wenck (1966: 14-21), McCawley (1968), Uemura (1997) and Akamatsu (2000).
Chapter 3. Phonology
will employ throughout this volume. Not shown are the vowel phonemes /a e i o u/, broadly realized [a e i o w]. Figure 3.2 contains 21 or 22 consonant phonemes, depending on the inclusion of /v/, which most Japanese speakers have great difficulty pronouncing. Vance ( 2008a: 81 ),for example, considers [v] a 'foreignism' and it is found only in the most innovative pronunciations, generally realized[~]: e.g. [~enetsia] 'Venice: The plosive phonemes /p b t d k/ are realised [p b t d k]. The voiced velar plosive /g/ is typically [g), although maybe [IJ] or [y] for some speakers, chieflyword-internally. 38 The affricates I c cj/ are realised [ts cs; J~]. The fricatives If s Sf are [cp s so); /h/ is [~] before /i y/, but [h) elsewhere. The phoneme /z/ is typically an affricate [dz] when word-initial or preceded by IN/ or /Q/, but a fricative [z] elsewhere. The nasals /m n/ are [m n], the flap /r/ is typically [c ], and the approximants /y w/ are [j UJ:]. The mora obstruent /Q/ is the initial segment of what is traditionally termed a geminate consonant and has a variety of realizations. The mora nasal /N/ also exhibits a variety of realizations, including [m n IJ Jl N]. For a thoroughgoing description ofboth these mora consonants, see Vance (2008a:96-112). Although most conservative speakers typically have the mora obstruent /Q/ only before the voiceless obstruents /p t k s/, many contemporary speakers now have /Q/ before any obstruent (e.g. /sutaQfu/ 'staff: /beQdo/ 'bed'). Some of the most innovative speakers even have !QI before sonorants (e.g. /aQraa/ 'Allall, /muQraa/ 'mullah'), thus casting doubt on the very appellation 'mora obstruent'.
bilabial plostve affricate fricative nasal flap approximant
b
p
f (v) m
alveo14r t d c z s n r
(alveolo) -pa14tal
ve14r k
c
glottal
g
moraic Q
j
h
~
N
y
w
Figure 3.2. The Japanese consonants: an analysis of the contemporary phonemic system
In order to make clear the transcriptive differences between the conservative (Figure 3.1) and the contemporaryphonemicization (Figure 3.2), the conservative examples in (10-13) are retranscribed as contemporary in (10'-13'). (10')
/fuhoo/ /hihoo/ /hyoohoN/
[cpwho:] [~iho:] [~jo:h6N:]
'illegality' 'sad tidings' 'specimen'
38. See Hibiya (1995, 1996, 1999), Saito (1997:87-89, 2003:12), Joo (1998:131-132) and Vance (2008a:214-222) for more detail.
73
74
Loanwords in Japanese
(11 ')
(12')
(13')
I tecu/ /hitozuma/ /toW /hanaji/
'iron' 'married woman' 'land' 'nosebleed'
[tetsw] [~itozwma]
[toq;i] [hanar~i]
/susi/ /jizeN/
[sw~i]
leal /kucija/
[c~a]
!sui
[~w]
/jo/
[J~o]
[J~izeN:]
[kwc~iJ~a]
'sushi' 'charity' 'tea' 'fresh tea leaves' 'type, kind' 'preface, foreward'
One of the most salient features of the Japanese language is its moraic structure. The mora 'functions as the unit oflength ... [and] the length of a phrase [is] roughly proportional to the number of moras it contains' (McCawley 1968: 131). There are three types of mora: (C)(G)V structures, where the optional onset Cis a consonant, the optional glide G an approximant, and V a vowel the mora nasal N the mora obstruent Q
The role of the mora in temporal regulation and phonological length, and also as a unit of segmentation and perception, is indisputable. Nevertheless, the syllable, as defined by Hattori (1954, [1958]1960), is also of consequence. One area where this is particularly so, as we shall see later in this chapter, is with suprasegmentals (§3.3) and mora-clipping (§3.4). A word can be divided by syllable or by mora, although the result may differ. This is illustrated in (14), where all the examples are gairaigo, a dot indicates a syllable boundary and a dash a mora boundary. A monosyllabic gairaigo may have anything from one to three moras (14a-c), and a bisyllabic gairaigo anything from two to six (14d-h). Monomoraic syllables are termed 'light' and are of the form (C) (G)V. Bimoraic syllables are 'heavy' and of the form (C)(G)VV, (C)(G)VQ or (C)(G) VN. Trimoraic syllables are 'lsuperheaV}lJ' and of the form (C)(G)VVN, (C)(G)VVQ or (C)(G)VNQ. 39 The gairaigo stratum is the only stratum where all three syllable types occur freely. Superheavy syllables never occur in the Sino-Japanese stratum
39· (C)(G)VNQ superheavy syllables are not found in pure gairaigo, though may appear in hybrid compounds (§1.2.5): e.g. roN.IdoNol.ko 'Londoner: 'Heavy' and 'superheavy' syllables may also be termed 'long' and 'extra-long: Issues relating to syllable division in the case of certain vowel combinations, as well as issues of accent placement, mean that the question of
Chapter 3. Phonology
and are rare in the native and mimetic strata. A syllable, whether light, heavy or super-heavy, must be either open or have a coda inN or Q. This rigid structure has important consequences for the adaptation of closed syllables and consonant clusters from donor languages, as will become clear in §3.2.
syllabic division moraic division
(14)
a. b. c. d. e. f. g. h.
'the (only/best)' 'pi, 1t' 'chain' '(comedy) gag' 'jet' 'kitchen' 'switcher (locomotive)' 'mainline'
za pai ~eeNI gya.gu jeQ.to kiQ.ciN lsuiol.caa lmeeNI.~
za pa-i ce-e-N gya-gu je-Q-to ki-Q-ci-N su-i-Q-ea-a me-e-N-ra-i-N
Table 3.1 shows the permissible mora types in both the conservative and contemporary phonemic systems, transcribed in contemporary fashion. The uppermost row shows the contemporary consonant phonemes, the leftmost column the vowel phonemes. Moras with glides (e.g. /bya/) and the two mora consonants I Q N/ are ignored. A blank cell indicates that a mora exists in neither system, a shaded cell indicates a mora found only in the contemporary system. All other cells represent a mora found in both systems. Bracketed moras are marginal. An alternative reading of Table 3.1 is that filled cells indicate the moras found in Japanese, while shaded cells indicate moras found only in gairaigo. Table 3.1. Permissible mora types in conservative and contemporary phonemic systems
p b t d a pa ba ta da pi bi ti di u pu bu tu du e pe be te de
c
k g c j f ka ga ca ca ja ta ki gi d ci ji fi ku gu cu cu ju fu
(v)
s
z
s
h
m
n
r
w
y
(va) sa za sa ha rna na ra wa ya (vi) (si) (zi) si hi mi ni ri (wi) (vu) su zu su yu mu nu ru
ke ge ce ce je fe (ve) se
~
Ae he me ne re (we) (ye)
o po bo to do ko go co co jo fo (vo) so
zo
so ho mo no ro (wo) yo
In a typical conservative phonemic system, such as that in Figure 3.1, the existence of the moras /wi we wo ye/ is moot (see the discussion in Vance 2008a: 90-92). In a contemporary system, the more innovative speaker may show the monomoraic articulations /we wi wo ye/, as opposed to bimoraic /ue ui uo ie/, in some gairaigo. Examples include /webu/ 'internet, the web: /winipegu/ 'Winnipeg: /wookumaN/
what constitutes a heavy or superheavy syllable is considerably more complex than described here. See Vance (2008a: 131-138) for a useful overview of the difficulties.
75
76
Loanwords in Japanese
'walkman' and /yeeru/ 'Yale'. Many innovative speakers also have a /\·\'/-glide, although generally only after ikl and, again, only in gairaigo. Such speakers thus have monomo.mic /kwa kwi kwe kwo/, as opposed to bimo.mic /kua kui kue kuo/. A lye! -glide (e.g./kyel) is typically not apparent. By no means all contemporary speakers employ all, or indeed even some, of these monomoraic articulations. The situation is extremely complex (see Vance & Matsugu 2008) and throughout this book my usage reflects the spelling most frequently found (§5.1). For many speakers, what I transcribe distinctively as /ue/ and /we/ are realised identically as bimoraic [me]. For others, they are realised distinctively as bimoraic [me] and monomoraic [ll{e], respectively. While a few innovative speakers have the moms /si zi/ in an extremely restricted set of words (see discussion prior to (28) on p. 84, the moras /hu yi wu/ are still not evident even in the most innovative contemporary pronunciations.
3.2
Adaptation
Borrowing routes in Japanese are schematized in Figure 3.3. The source (stage 0) of a donor word may be auditory or, more likely; orthographic. If it is the latter, it is frequently assigned a dictionary pronunciation (stage$). As stated in (3) on p. 10, my definition ofgairaigo includes the condition that a loamwlrd must have 'undergone adaptation' to Japanese phonology. In the case of an auditory source, adaptation (stage 8) is based on the auditory input; in the case of an orthographic source, it is based on dictionary traditions. These three different routes result (stage 0) in three different kinds ofloan: auditory; dictionary and spelling. Of these three, the auditory loan is the least common. Borrowing in Japanese has occurred in a distant setting with little direct auditory contact (§1.1). This \·\'as especially true for all but the earliest Dutch borrowings (§2.3) and has also been very much the case in the modern era (§2.4). Japanese has been the dominant language in a state which currently has no land borders. Other languages indigenous to the archipelago have had comparatively tew speakers. There has never been a sufficient number of speakers of non-indigenous languages resident in Japan f.ix significant auditory contact to occur. With the exception of Sakhalin (§2.4.1) and the brief US occupation of all Japan from 1945-1952 and of Okina\'\'a until1972, the number of speakers of European languages residing in Japan has generally been small. 40 Although the number of Korean· and Chinese·speaking residents
40. Based on nationality statistics of foreign residents (STK 2008), in 2005 this amounted to approximately 515,000 people or just 0.4% of the total population of Japan. Histo.ti.cally, this figure is at an ail-time high, boosted by an infltLx from Brazil. By comparis..-:>n, the figure for 1985 is estimated at onJy 58,000 or a mere 0.04% of the then population of Japan. Neither of
Chapter 3. Phonology
has been higher (§2.5), until recently Korean carried little prestige. In a recent government survey (BBK 2008), 59% of respondents said they had not been spoken to by a foreigner in Japan at all in the previous two years. Of these, 12% stated they would not even attempt to reply to one if they were.
0
SOURCE
Auditory
Orthographic
II
J, r········Di-~ti·~~~;y·········i
i
i
pronunciation assigned
i
i
l----------------------------------~
•
ADAPTATION
MSED ON•..
J, Auditory input
II
Dictionary traditions
•
RESULT
Figure 3.3. Borrowing routes in Japanese
these figures include naturalized speakers of European languages, whose numbers are very small. A variety of English, as well as an English-Japanese mixed language, have been spoken on the Bonin (Ogasawara) Islands since the early 19th century, but the number of speakers has numbered only in the low hundreds (Long 2007).
77
78
Loanwords in Japanese
The early decades of the Western borrowing phase (§2.4) did witness a more significant proportion of loans likely the product of auditory contact, especially those whose source was a 'port lingo' (Daniels 1948; Inoue 2007; see §2.4.4). It was, however, Japanese writers and essayists of the period who were chiefly responsible for introducing loanwords into the general speech community. Many of these, a prime example being the novelist, poet and doctor Mori OgaL~l,:11- (1862---1922), came from upper--class backgrounds, had contact with non-Japanese, travelled overseas and were the 'digesters, absorbers and propagators of a wide range of Western knowledge' (Yamada 2005: 82). Many would lard their works with \·\'hat at this stage were widely unintelligible gaikokugo. Those gaikokugo which underwent adaptation followed a variety of pathways depending on the author; the Japanese spellings employed to write them were often subjective, highly unstable and exhibited \·vide variation (§5.2). In time, as many of these words became established in the gene.ml speech community, levels of comprehension increased and what were gaikokugo became gairaigo (J--4, p. 10). By around the 1920s, orthography and pronunciation too had in most cases stabilized and the forms found today begin to be encountered (opus cit.: 93-95). As the influence of foreign -language, especially English-language, education grew, so awareness of correct donor pronunciation heightened and the probability of a more accurate adaptation increased. By the post-war period, it was no lon .. ger a privileged few authors who disseminated western knowledge and loanwords among the now more educated and increasingly sophisticated Japanese speech community, but academics, teachers, journalists, translators and eventually televi .. sion presenters and internet bloggers. Even now, though, the teaching of foreign languages in Japan is still a profession dominated by monolingual Japanese speakers, relying heavily on gmmmar/translation methods, and placing a great deal more emphasis on reading, writing and spelling than on perfecting pronunciation (Gottlieb 2005:31-32; Mantero & lwai 2005; Daulton 2008:2-3). This, coupled \·vith lack of auditory contact, means that the influence of donor orthogmphy on any potential loanword disseminator remains strong. Loans with an orthographic source have thus come to comprise the bulk of the gairaigo stratum and, for the most part, the f(Hmation of dictionary traditions (stage 8 in Hgure 3.3) have been responsible for their increasingly accurate adaptation. Although dictionary traditions vary according to donor language, all have in common the fact that their adaptation rules were established and stan· dardized by Japanese scholars of foreign languages, then perpetuated through their pedagogical practices and foreign language textbooks. Since its spelling is notoriously opaque, donor words from English are typically assigned a dictionary pronunciation at a point prior to adaptation (stage 8). French spelling too is opaque and a dictionary pronunciation is t)'lJically assigned here also. For major
Chapter 3. Phonology
donor languages \·\'hose spelling is more transparent, such as German, Russian or Italian, the assignment of a dictionary pronunciation is less important. A borrow .. ing whose source is orthographic (stage 0), which has been assigned a dictionary pronunciation (stage 0), and which has undergone adaptation based on a dictionary tradition (stage 0) is a dktionary loan. Dictionary traditions are, in effect, prescribed adaptation strategies. Since, however, each donor language has its own prescription, the same source sound may follow more than one adaptation pathway. Examples include the substitution of 1J borrowed from Chinese and Korean on the one hand, and Western languages on the other (68 ..·70); the substitution of schwa a, which varies substantially from donor language to donor language (116-120); and the varying epenthetic vowels used after German, Dutch and Russian donor x (127-131). Perhaps the most striking example is donor v. In gaimigo borrowed from English and French v has been systematically substituted by Japanese fbi, while in those borrowed from German and Russian, it has typically been substituted by iw u 0/ (31 ..·35). Given their nature, dictionary traditions are in many cases grounded in .f()reign language scholars' deep knowledge of donor language phonemics: see for example the case of hench word-final n (60-62). As noted by Quackenbush (1974:61), when many dictionary traditions were being formulated in the late 19th century, English language scholars were heavily influenced by fi)reign advisors (oyatoi gaikokujin: cf. §2.4.4) from Britain. It is probably .f()r this reason that adaptation patterns f(H, amongst others, the vowel found in the word lot (89 ..90), as well as the English rhotic vowels (111 115), mirror English English and not American English pronunciation. 1hese dictionary traditions based on English English still hold, despite American English being taught exclusively in secondary and ovenvhelmingly in tertiary education. Loans whose source is orthographic also include a not insignif.icant number of cases where a dictionary pronunciation has not been assigned at stage 0. Here, when adaptation (stage 0) has been based on a spelling which is an inaccurate representation of pronunciation, the result is a spelling loan (tsuzuriji hatsuon ~ ~ ::f 9&"!}, lit. 'spelling pronunciation: in the Japanese tradition). 41 'This is often the case when a donor word is a neologism which does not appear in a dictionary. A recent example is wikipedia 'Wikipedia' for expected *wikipidia. Other, longer established, examples are listed in (15a). In each example the same donor word has spawned
41. When a dictionary pronunciation has not been assigned and the adaptation has been based on a spelling which is an accurate representation of pronunciation, then a spelling loan is indistinguishable from a dictionary loan. 'Ihis is not unusnaJ when the donor word is from a language with a highly transparent spelling system, such as German, Russian or Italian.
79
So
Loanwords in Japanese
multiple loans. The modern standard form is a spelling loan and highlighted in bold. Other attested forms are either non-standard or obsolete. These include dictionary loans (underlined), auditory loans (italicized) and other non -standard or obsolete spelling loans. With still other spelling loans, earlier variants are unattested and whether a dictionary or auditory loan ever existed is unknown. Examples are shown in (l5b), with the expected, though unattested, dictionary loan asterisked. (15)
a.
[0.~1872] ObU.N, (")1919] OOVUN, [0 1953] OObUN Eng. oven [01890] kopii, [0 1890] kQQ,P.jj Eng. copy Eng. hockey ro1907] hoo.kii, [0 1931] hoQkee Eng. studio [0 1925] sutajio, [01927] sutyudio, [0 1936] sutadio
b.
Eng. radio
rajio, *reejio
Eng. buzzer Eng. chocolate Eng. news Eng. sponge Eng. monkey Eng. condom Eng. zero
buzaa, *P.I!Y.ll!
cokoreeto, ·~QkQr.~Q.tQ D}'UUSU,*~
supoNji, *!i.UJ?.l!!~lj!
moNkii, "Dlf!.Nk.U. koNdoomu, "koNdomu zero, *jiiro
'oven '(photo)copy' 'hockey' 'studio'
'radio' 'buzzer' 'chocolate' 'news (programme)' 'sponge' 'monkey' 'condom' 'zero'
In this section, I take a purely descriptive, theory-neutral stance in outlining adaptation 42 processes (stage C) in Figure 3.3). Moreover, my description is essentially a synchronic one, although reference will be made to redundant adaptation pathways which have had important consequences fix modern gairaigo phonology; as well as to pathways which have seen noteworthy diachronic change. Such pathways should not be overlooked, as has sometimes been the case in theorybased analyses. While most theoretical models presume the 'foreign input' to be auditory; I presume it to be a dictionary tradition, unless otherwise noted. Recent research by Vendelin & Peperkamp (2006) and Tranter (2011) on orthography and loanword adaptation processes in general is therefore most welcome. ·when foreign words contain sounds or structure that is not found in the native language, the foreign input is mapped to the 'most similar' structure that is available in the native language ... [M]uch has been written about the nature of the input language representation namely, whether the phonological or phonetic representation oft he lending language serves as an input to the adaptation - and substantial arguments have accumulated on both sides. Kang & Rice (2008: 273-274)
42. Adaptation may also be referred to as 'nativization' or 'accommodation: although the three terms may exhibit subtle differences in meaning depending on theoretical stance.
Chapter 3. Phonology
Such arguments go back to at least Polivanov (1931). In line with my nontheoretical stance, I commit here to neither the phonological (e.g. LaCharite & Paradis 2005), the phonetic (e.g. Silverman 1.992), nor any other camp. A.ll stage 8 forms are indicated neither phonemically, as /mputi, nor phonetically, as [rnput], but rather between vertical bars, as lrnputl. The ensuing discussion f(Kuses exclusively on the adaptation patterns found in standard Japanese non·· standard varieties may, as noted by Tinnouchi (2007: 36) and others, exhibit pat·· terns different from those described below. Throughout this chapter, the symbol ~ is used to indicate 'adapted to' or 'substituted bY, while ~ indicates 'adapted from' or 'substituted from'. As Japanese is phonemically more impoverished than all its main gairaigo donor languages, as \·\'ell as suffering from strict syllable constraints (14), adaptation can be extreme. The language has two major adaptation strategies: phonic substitution and epenthesis. Phonic substitution (§3.2.1) refers to the strategy in which a speaker attempts to replace a sound in a donor language with one in her native language: see Paul (1880) or Bloomfield (1933:445ff) for early work. The second major adaptation strategy, epenthesis (§3.2.2), is the insertion of additional phonemes by the borrowing language. In the case of Japanese this can be further subdivided into vowel epenthesis (§3.2.2.1.) and epenthesis of the mora obstruent /Q/ (§3.2.2.2). There is, in addition, a third minor strategy of adaptation, deletion, confined almost entirely to auditory loans. This I deal with in §3.2.3. 3.2.1
Phonic substitution
The substitution of donor consonant phones is examined in §3.2.1.1, of vowel phones in §3.2.1.2. 3.2.1.1
Consonants
Consonant substitution will be described according to manner of articulation in the following order: plosives (16-27); fricatives (28-48); affricates (49-55); nasals (56-72); trills, taps, flaps and laterals (73-75); approximants (76-85). As I am unaware of any donor words containing them, I do not consider the plosives [l q ?], fricatives [cp j'n ~ fi. ~ ~ H 1·], nasals [ll) n.N], trills, taps, .flaps and
43· The phenomenon known in Spanish phonology as yefsmo means that the two phonemes I !.I and /j/have merged in many dialects (Penny 2000:120-121, 147-148).As (74) and other examples attest, the donor phone in most borrowings into Japanese would appear to be 1..<1. Some very recent names, however, exhibit the substitution UJ ~ /j/. These include members of the victorious Spanish 2010 PIPA Wo.rld Cup squad, such as (David) Villa(~ /bija/), and Spanish football dubs such as Sevilla PC (~ /sebiija/). Compare the place name Seville in (30).
81
82
Loanwords in Japanese
t
laterals [B v c r l 44 l~ L 1], or approximants [v ·l uy]. Nor, for the same reason, do I consider any non·pulmonic consonants (clicks, implosives and ejectives). Japanese has a broadly similar array of plosives to most European donor lan·· guages and, with the exception of It dl before close vowels, donor IP b t d k gl ~ ip b t d k gi, as illustrated in (16). Though examples are few, English intervocalic It dl may~ iri if the source is auditory (§3.2.3). Word··finalltl from Korean typi·· cally~ iQ!, for which see (171e). (16)
lrl ~ipi lbl ~/b/ ltl ~it/ ldl ~idi lkl ~ /k! lgl~tw
Eng. purple Eng. bubble Eng. tartan Eng. dodo Eng. cookie Eng. goggles
paapuru baburu taataN doodoo kuQkii googuru
'purple' 'bubble (economy)' 'tartan' 'dodo' 'cookie, biscuit' 'goggles'
Since, in the conservative phonemic system, /t di have affricate or fricative allophones before close vowels (11, 11'), donor It dl before a substituted close vowel ~ /cu zu ci ji/ in the oldest gairaigo. Some examples are given in (17). Th.ese have no contemporary doublets pronounced *tuberukurin, *duQku, *tifusu, *digitarisu. With the most recent borrowings, adaptation is strongly contemporary, as shown in (18). Here, conservative doublets such as tacuu, dijerizuu, Ciramisu andjiibuijii are less widespread. There is a large intermediate layer, illustrated in (19), consist.· ing of conservative and contemporary doublets. The conservative pronunciation is now generally confined to older age cohorts (Sakamoto 2002:57 -59), lower socioeconomic groups, and rural or peripheral dialects. Crawford (2008:60) has found that the contemporary pattern seems to have appeared earlier with donor words ending in Itil, ldil, ltrl) I and ldriJ I· (17)
(18)
a. b. c. d.
Ger. Tuberkulin Du. doek Du. typhus Du. digitalis
jigitarisu
a.
Eng. tattoo .Eng. didgeridoo It. tiramisu Eng.DVD
tatuu dijeriduu tiramisu diibuidii
b. c. d.
cuberukuriN ZUQkU
cifusu
'tuberculin' 'canvas, sackcloth' 'typhus' 'digitalis' 'tattoo' 'didgeridoo' 'tiramisu' 'DVD'
44· The lateral fti.cative [t] --7 /r/ in a few borrowed Welsh names. Since [l] --7 /sur/ would be the most likely substitutional pathway for both dictionary and auditory loans (e.g. Wel. Llanelli --7 • /suranesurii/, rather than the attested /ranerii/), it is highly likely these names have been borrowed via English and an anglicized 111.
Chapter 3. Phonology (19)
a. b. c.
d.
Fr. 'l'oulouse Ger./Eng. Himlu Eng. team .Eng. dilemma
cuuruuzu ~ tuuruuzu hiNZUU ~ hiNdttU ciimu ~ tiimu jireNma ~ direNma
'Toulouse' 'Hindu' 'team' 'dilemma:'
An exception to the adaptation pathways in (17 19) is (20). The vowel of donor ldil or ld1l may be lowered during adaptation, to give /de/. Although by no means universal (note 18b, 19d), such lowering has occurred in many gairaigo, both new and old. An analysis by Crawford (2007) indicates that the proportion of English loanwords subject to this /del-lowering pathway has remained relatively stable over time, and been generally confined to donor word-initial position (20a-c). Some loans may have doublets in /di/ (20cd). The reason for the substitution /de/ is unclear. As the spellings of (20c) digital or (20d) candy show, these are not spelling loans (§3.2). Further, it is difficult to explain why the phenomenon's voiceless coun·· terpart should occur with considerably less frequency. Donor ltil and Itil are lowered to /tel with only extreme rarity: e.g. suteQki \\'alking stick, cane' f- Eng. stick; suteQkaa 'sticker' f- Eng. sticker. It can only be concluded that lowering to /de te/ is an old dictionary tradition, dating from before the birth of the contemporary phonemic system, whose motivation was based on preserving donor It dl as plosives. (20)
a. b. c. d.
[nia] [nia] [nia - +1960] [ 0 1904 ~ 0 1952]
Eng. decode Eng. decanter Eng. digital Eng. candy
dekoodo dekaNtaa dejitaru ~ d.ijitaru kyaNdee ~ kyaNdii
'decode' 'decanter, carafe' 'digital' 'sweets, candy'
Aspiration in Japanese is non-phonemic and ignored during substitution. This is the case both for borrowings from donor languages where aspiration is non-phonemic (the majority), as well as those from donor languages where it is phonemic. As an example of the former case, the French town Pau, with syllableinitial unaspirated lpl, and the American writer (Edgar Allan) Poe, with syllable-· initial aspirated IPhl, both ~ poo. ·with borrowings from the only hvo donor languages of any significance \·\'here aspiration is phonemic, Chinese and Korean, both aspirated (21) and unaspirated (22) voiceless plosives typically~ voiceless plosives in Japanese. The Korean unaspirated phones IP t kl undergo predictable non-phonemic voicing to lb d gl when they occur between two voiced phones and these typically~ voiced in Japanese (23). Korean, in fact, has a three-way contrast IP ph P"l in its plosives and alTricates, 45 termed 'lax .. aspirated .. reinforced'
45· 'Ihe transcription IP* I is that used by Ladefoged & Maddieson (1996: 55-57) to mark what they describe as the 'stiff voice' (a slight degree of laryngealization) of the reinforced (what they term the fmtis) series.
83
84
Loanwords in Japanese
by Lee & Ramsey (2000: 61-63). Though examples are scarce, the reinforced plosives typically~ voiceless in Japanese, as shown in (24). (21)
(22)
(23)
(24)
IPhi IPhi lthl lthl
BeL pai Kor. p'yongyang Cant. wan4 tan 1 Kor. t'aekwondo
IPI IPI ltl lti lkl lkl
BeL xiaol6ngbao Kor. pulgogi BeL douMnjiang Kor. taegu Bei. gaoliang Kor. kimch'i
lgl lbl
Kor. han'gul Kor. kalbi
IP*I
h~-
pai pyoNyaN waNt aN tekoNdoo
j~()J
~* Ell 'els:. /j\~{r¥; ~Jl7]
.:R.:j;,R~
c.Jl-1~:;*
~~l
~~ ~[l:l]
haN guru karubi
¥,?3[
Kor. ppongtchak
sooroNpoo purukogi toobaNjaN tegu kooryaN kimuci
'mahjong tile' 'Pyongyang' 'wanton dumplings' 'taekwondo' 'steamed dumpling' 'barbecued beef dish' 'chilli bean sauce' , 'Daegu 'type of millet' 'pickled cabbage'
'Korean alphabet, language' 'thinly sliced rib meat'
poNcaQku
'genre of Korean pop music'
The palatallc Jl and uvular lq Gl plosives, as well as the glottal stop I? I, occur only in a very small number of borrowings, mostly names. Here, lc Jl ~ /c j/ (25), lq Gl ~/kg/ (26), and I? I~ 0 (27). Although the question of whether some of the words in (25-27) were borrowed directly or via an intermediary is open to question, the outcome in terms of substitution would in all likelihood be the same. lei IJI
Cz. Navratil.ovci
(26)
lql IGI
Ar.lqur?a:nl Far.IGoml
(27)
l?l
Haw. O'ahu lo?nhul
(25)
naburacirowa naji
Hu.Nagy
'(Martina) Navratilova' 'Nagy (Irnre)'
kooraN - kuruaaN gomu oafu
'Koran/ Quran' 'Qom'
'Oahu'
,-._,
The alveolar fricatives Is zl ~Is zi (28), except when they precede a substituted /i/, in which case they~ Is j/ (29). Even amongst the most innovative Japanese speakers the moras /si/ and /zi/ are rarely found. Perhaps the only examples are the occasional use by sports commentators of siizuN for Eng. season (Ishiwata 2001: 36) and the very careful articulation of the English letter Cas sii or the North American letter Z as zii. These three gairaigo are more commonly siizuN, .5ii and jii, respectively. The English plural morpheme -s, when pronounced lzl, frequently~ Is!. Examples of this dictionary tradition includ.e seerusu 'sales: kabusu '(Chicago) Cubs' and arisusupuriNgusu 'Alice Springs'. (28)
lsi lzl
Eng. save Ger. Seil
seebu zairu
'save' 'climbing rope'
Chapter 3. Phonology (29)
lsi lzl
.Eng. scene Eng. magazine
siiN magajiN
'scene' 'magazine'
The voiced bilabial fricative 1~1 ~./b./:
(30)
Sp. Sevilla Sp. C6rdoba
sebiria korudoba
'Seville' 'Cordoba'
'Ihe adaptation of the voiced labiodental fricative lvl is complex. In borrowings from Russian and German, lvl typically~ ivd or lui before substituted /a e i oi (3la···d). There are also cases where lvl ~fbi (31ef) or even~ 0 (3lg) before the same four substituted vowels. Before a substituted or epenthetic /u/, however, lvl typically ~ 0 in borrowings from Russian (32a), but ~ !b! in those from German (32b). In one borrowing from Czech, lvl ~ !bi (32c).
(31)
a. b. c.
d. e. f
g. (32)
a.
b.
Ru. MocKBa lmoskval Ger. Wedeln Ger. Wien Ru. Bop;Ka lvotkal Ru. 6o!IbiiieBMKU lboBeviildil Ger. Wolfsburg Ru. BnanuBOCTOK lvlad.iivostokl
mosukuwa ueederuN wiiN llOQka " WOQka borusebiki borufusu buruku urajiosutoku
'Moscow' 'wedelling' 'Vienna' 'vodka' 'Bolshevik' 'Wolfsburg' 'Vladivostok'
Ru. Bnani-mocTOK lvladiivostokl urajiosutoku 'Vladivostok' Ger. Wuppertal buQpaataaru 'Wuppertal' 'Vltava, .Moldau' Cz. Vlatava burutawa
In borrowings from English, French and Italian, lvl regularly ~ /b/ (33). A few innovative speakers have !vi, as the doublets in (34) show.
(33)
Eng. veteran Fr. vinyle It. da Vinci
(34)
Eng. violin Fr. deja vu It. Venezia
beteraN biniiru dabiNCi
'old hand' 'soft plastic, vinyl' '(Leonardo) da Vinci'
baioriN ~ vaioriN dejabyu ~· dejavyu beneCia -- venecia
'violin'
'deja vu' 'Venice'
Some innovative doublets with /v/ from Russian and German sources are shm·\'n in (35), but these are less common than from English, French or Italian. Long-established place names such as (31ac) are rarely articulated mosukuva or viiN. (35)
Ru. lifBaH Iivan I Ru. HoBocM61ipc:K lnovosiibiirskl Ger. WeifJbier Ger. Wo[fsburg
iwaN- ivaN nobosibirusulm- novosibirusuku waisubiiru - vaisubiiru borufusuburuku - vorufusuburuku
'Ivan (the 1e1Tible, etc.)' 'Novosibirsk' 'weiss bier' \ 1•lolfsburg'
This difterence between the treatment of donor Ivi from German and Russian on the one hand, and English, French and Italian on the other, is due to dictionary
85
86
Loanwords in Japanese
traditions intended to preserve phonemic contrast in the latter three languages. As will be described later in (77---79, 139---141), both English and French donor lwl -4/w u 0/. Thus, lvl-4/b/ to preserve the contrast. German and Russian have no lwl and thus lvl is free to be substituted as the acoustically closer /w/ or /u/, preferred over /b/. Early German and Russian textbooks corroborate this. Por example, Okazawa (1946: 12), a Russian textbook for Japanese beginners, describes the pronunciation of the Russian letters
[0 1906] [ 0 1910] [*1913] [*1920]
Eng.fault Eng.fence Eng.film Eng.fan
(37)
[*1615] [0 1837]
Du. koffie Du. morfine
(38)
[0 *1713] [+1832]
(39)
Du. kamfer Du.flanel
fooruto feNSU firumu faN
'fault (in tennis, etc.)' 'fence' '(camera) film' 'fan'
koohii moruhine kaNfuru fura(N)neru
'coffee' 'morphine' 'camphor' 'tlannel'
a. Er1g. fuse hyuuzu "' fyuuzu 'fuse' b. Eng. platform (puraQto )hoomu- puraQtofoomu '(station
TI1e dental fricatives they -4 Is jl (41).
M
computer) platform'
10 ol -4 Is z/ (40), except bef()re a substituted /i/, where
Chapter 3. Phonology
(40) (41)
IOI Ia I IEll lol
'marathon~
.Eng. marathon Eng. leather
marasoN rezaa
'leather'
Eng. theatre Eng. smoothie
siataa sumuujii
'theatre' 'smoothie'
The post-alveolar If 31 and retroflex I~ '4.1 fricatives ~ is j!, as shovm in (42). Since the conservative phonemic system does not admit the mora /se/, there are several earlier borrowings where ~~ ~ Is/. These are shown in (43), along with their more frequent contemporary doublets. (42)
IJI 1~1
131 l1;.l (43)
IJI IJI IJI
Eng. shampoo Bei. shimghdi .L~ Fr. genre Ru. Epe)lmeB lbriezn!ef]
saNpuu saNhai jaNru burejinefu
'shampoo' 'Shanghai' 'type, genre' '(Leonid) Brezhnev'
Eng. shepherd sepaado- sepaado 'shepherd dog' Eng. milkshake mirukuseeki ~· mirukukeki 'milkshake' .Eng. shade seedo - kedo 'sunshade, awning'
The voiceless palatal 1~1, velar lxl, uvular lxl, pharyngeal lhl and glottal lhl fricatives all~ /h/ (44), except before a substituted or epenthetic lui, where they ~ iff (45). A different pattern is observed in some older gaimigo borrowed from Dutch. Here, word-initiallxl ~lsi (46a), while the cluster lsxl ~ /suki (46b). More recent borrowings from Dutch may (46c), or may not (46d), follow this pattern. (44)
1<;:1 lxl lxl lxl Ihi Ihi Ihi
(45) (46)
Ger. Miirchen Ru. OxoTcK loxotskl Ger. Bach Heb. lxanuka:l Ar.lhama:sl Du. ham Eng. hormr
Ihi
Eng. hood
lxl
Sp./uan
a. b. c. d.
lxl lsxl lsxl lsxl
meruheN ohoocuku baQha hanukaa hamaasu hamu horaa
fuudo fuaN
Du. chemie
Du.sclwp Du. Schiphol Du. Enschede
'fairy tale' '(Sea of) Okhotsk' '(Johann Sebastian) Bach' 'Hanukkah' 'Ham as' 'ham' 'horror (movie)'
'hood' '(King) Juan (Carlos)' seemi sukoQpu sukipooru eNsuheede
'chemistry' 'shovel' 'Schiphol (airport)' 'Enschede'
The voiced uvular fricative IT! I found in most varieties of French~ /r/: (47)
Fr. ratatouille Fr. Paris
ratatuiyu pari
'ratatouille' 'Paris'
87
88
Loanwords in Japanese
As illustrated in (48a---d), the voiced velar fi-icative IYI ~ igi or /h/. Both pat-terns may be found in the same loan (48e). Befi)fe a substituted or epenthetic /u/, IYI ~iff (48f; also 129). (48)
a. b. c.
d. e. f.
Du. orgel 'organ' Du. Den Haag Du. Ghent Du. Nijmegen Du. van Gogh Du. Guliit
'music box' 'The Hague' 'Ghent' 'Nijmegen' '(Vincent) van Gogh' '(Ruud) Gullit'
orugooru haagu heNto naimeeheN baNgoQho furiQto
The voiceless alveolar affricate itsl typically~ ic! (49a---b), although in older borrowings it frequently~ /C/ beklre a following substituted /i/ (49c). Its voiced counterpart ldzl typically~ /z./ (50). Ger. Zeppelin Ru. Erron;HH ljelcinl Ger. Vakzin
(49)
a. b. c.
(SO)
Eng. Leeds Eng. AIDS
riizu eezu
ceQperiN ericiN wakuciN
' zeppelin' '(Boris) Yeltsin' ' • > vaccme
'Leeds'
'AIDS'
The voiceless postalveolar affricate ltfl ~ /C/ (51), while its voiced counterpart ld3l ~ /j/ (52). Since the conservative phonemic system does not admit the moras /ce je/ (cf. 43), there are a small number of early borrowings where l~f d3l ~ /s z/. Of the examples listed in (53), zerii, zeNtorumaN, mazeNta and sera have the contemporary doublets jerii, jeNtorumaN, majeNta and cera. (51)
Eng. c1wt room It. focaccia
(52)
Eng. agent Eng. jogging
(53)
[•1876] [ 0 1878] [ 0 1882] [•1901] [ 0 1907] [ 0 1909] [•1910]
CaQtoruumu fokaQca eejeNto jogiNgu
Eng. jelly Eng. gentleman Eng. gelatin Eng. magenta Eng. general strike
It. celio Eng. geranium
'chat room' 'focaccia'
'agent' 'jogging' zerii 7-eNtorumaN zeraCiN mazeNta 1.enerarusutoraiki sero zeraniumu
'jelly' 'gentleman' 'gelatin' 'magenta:' 'general strike-' 'cello' 'geranium'
Both aspirated (54a) and unaspirated (54b) donor affricates from Korean typivoiceless in Japanese. Korean reinforced aftricates also ~voiceless (54c ).
cally~
Chapter 3. Phonology 89 'Ihe unaspirated affricate la.l undergoes predictable non-phonemic voicing to U between two voiced phones and typically 4/j/ in Japanese (54d). While Chinese aspirated affricates also typically 4 voiceless (54e), unaspirated affricates often 4 voiced (54fg). (54)
a. b. c. d. e. f.
g.
lahl···-> /c/ Kor. chuch'e Kor. chuch'e lwl---~ ict Kor. tchigae 1.1 ····) lji Kor. chijim lt~hl ----) /c/ Bei. chiiofan lt~l--7 /j/ Bei. zhajiangmian Ia! ----) tjt Bei. majiimg
Ia! --7/ct
-'r-"11 -T''-"11 -~l7ll ~:.::-1~
fili11(
tt-loo JJf:~
CUCe 'juche' CUCe 'juche' clge 'spicy stew' cljimi 'pancake, jeon' 'fried rice' caahaN jaajaNmeN - jaajaameN 'fried sauce noodles' maajaN 'mahjong'
The voiceless bilabial-labiodental affricate lpfl from German etyma typically 4 /puf/, but occasionally /p/ or /f/: (55)
pufenihi ~- penihi pufarucu ~ farucu
Ger. Pfennig Ger. Pfalz
'pfennig' 'Pfalz'
'The alveolar nasal lnl 4 /n/, when donor word--initial or before a vowel (56). When donor word-final or before a consonant, lnl4 /N/ (57). A dictionary tradition dictates that orthographic (nn) often 4 /Nn/ (58). However, there are numerous cases where this tradition does not apply. Some are shown in (59). Besides the letter N, which 4 enu and not "eN, a major exception to (57) is word-final lnl in words donated from French. These do not 4 /N/, but 4 in/ (60) or /Nn/ (61), depending on spelling. Both are f()llowed by an epenthetic lui (§3.2.2.1). (56)
Eng. 11eeds Eng. unique
niizu yuniiku
'needs' 'unique'
(57)
Eng. iemon Eng. panther
remoN paNsaa
'lemon' 'panther'
(58)
Eng. thinner Eng. savannah
SiNnaa sabaNna
(59)
Eng. scarmer Fr. mayonnaise
sukyanaa mayoneezu
(60)
Fr. cuisine Fr. Seine
(61)
Fr. Cannes Fr. Cezanne
kyuijiinu seenu kaNnu sezaNnu
'paint thinner' 'savannah' 'scanner' 'mayonnaise'
'cuisine' 'Seine' 'Cannes' 'Cezanne'
90
Loanwords in Japanese
Peperkamp et al. (2008) explain (60-61) as a consequence of the 'strong release' of word--finallnl in .French. This is the so-called 'e muet' or 'e caduc: an epenthetic schwa [d] appearing in modern French 'most often after liquids and nasals' (Coveney 2001: 182) and especially frequent in careful pronunciation. Another explanation, which Peperkamp et al. (opus cit.: 136) reject, is that it is a dictionary tradition intended to preserve the phonemic contrast in French. As we shall later in (109···110), a donor nasal vowel lVI ~ !VN/. Thus, word--finallnl ~ /Ni, as in (57), would lead to a loss of contrast, since both word-final lVI and word-finaljVnl would ~ /VN/. The patterns in (60-61), unique to French, allow contrast to be preserved: (62)
Fr. Caen Fr. Carmes
IVI--t /VN/ IVnl --t /VNn/
kaN kaNnu
'Caen' 'Cannes'
TI1e bilabial nasallml--t /m/ (63), except when followed by a bilabial or labiodental where it typically--tiN! (64). As with (nn) (58), a well--established dictionary tradition has it that orthographic (mm) often~ /Nm/ (65). Some cases where this dictionary tradition does not apply are shown in (66). When (mm) is generated by the English morpheme -ing, the tradition never applies. Compare (67) to (65b). , 'mummy (63) Eng. marna mama 'jam' Eng. jam jamu , 'hamster hamusutaa Eng. hamster (64)
Eng. symbol Du. kamfer
(65)
a. Eng. hammer b. Eng. ammonia
(66)
Eng. Hummer Eng. backgammon
(67)
Eng. jamming Eng. swimming
siNboru kaNfuru
<
symbol' 'camphor'
haNmaa aNmonia
'hammer' 'ammonia:'
hamaa baQkugyamoN jamiNgu suimiNgu
'Hummer' 'backgammon'
'jamming' 'swimming'
Dictionary traditions mean that the velar nasal ltJ I exhibits differing patterns of substitution depending on donor language. With Korean and Chinese borrow·· ings ltJI ~ /N/ (68). With European donor languages, llJI typically~ /N/ only when followed by a velar or alveolar plosive (69), otherwise IIJI ~ iNgi (70). There are a small number of mainly auditory loans from English where IN! is found for expected /Ng/, f.ix which see (176). (68)
BeL zh4iiangmirln f'F'tt- w jaajaNmeN 'fried sauce noodles' Kor. pyongyang pyoNyaN 'Pyongyang' ~~t .iff: !f.} maajaN 'mahjong' Bei. majiang
Chapter 3. Phonology
(69)
Eng. pink Eng. hu11ger strike Eng. Vvashington Eng. Hunti11gdon
piNku haN gaasu toraiki wasiNtON haNtiNdON
'pink' 'hunger strike' 'Washington 'Huntingdon
(70)
Ger. Doppelganger
doQperugeNgaa herusiNgu bori gyaNgusutaa raNkiNgu
'doppelganger' 'Helsingborg' 'gangster' 'ranking'
Sw.Helsingborg Eng. gangster Eng. ranki11g
The palatal nasal !P-I ~ /ny/ or /niyi before a substituted or epenthetic /a o u/ (71), but~ in/before /e i/ (72). (71)
It. lasagna
razaanya - razaniya eruniinyo buuroonyu
Sp. elNiito Fr. Boulogne (72)
It. bologrzese
boroneeze bobini
Fr. Bobiguy
'lasagne' 'el Nifio' 'Boulogne'
'bolognaise (sauce)' 'Bobigny'
The few donor words containing a prenasalized consonant are mainly names from Bantu languages. Although prenasalized alveolars typically~ iNC! (e.g. Ndebere 'Ndebele'), prenasalized bilabials typically~ /muC/ (e.g. mubeki '(Thabo) Mbeki').
The fact that Japanese has only one liquid, ir/, results in all trills, taps, flaps and laterals from donor languages~ /r/:
Ger.Ruhr Ru. HopMa !normal Eng. lock
ruuru noruma roQku
'Ruhr' 'quota' 'lock'
The palatallateraliA.I ~/riy/ or /ry/ before a substituted /a o ui (74), but~ /r/ before /e i/ (75). Examples are scarce. (74)
Sp. paella
Sp.Llosa (75)
It. Gigli
paeriya ryosa jiiri
'(Beniamino) Gigli'
The alveolar approximant English varieties, ~ /r/: (76)
Eng. rule Eng. pride
'paella' '(Mario Vargas) Llosa:'
ruuru puraido
111, the rho tic consonant found in the majority of 'rule' 'pride'
91
92
Loanwords in Japanese
~
The voiced labio--velar approximant lwl ~ lwl before a substituted /ai (77a), f<J beklre a substituted /u/ (77b), and~ either /u/ or /w/ elsewhere (77c-e).
(77)
Eng. wine Cant. wan4 tan 1 ~;fi;.
waiN waNtaN
b.
Eng. Worcester sauce Pol. L6di lwuaj
usutaasoosu uQCi
c.
Eng. Wichita Eng. weekend
d.
Eng. Wales Eng. waiter
e.
Kor. wl5n ·@ Eng. water polo
a.
wiCita wiikueNdo weeruzu ueetaa
'wine'
'wonton dumplings' 'Worcestershire sauce' 'L6di'
'Wichita' 'weekend'
'Wales' 'waiter'
WON
uootaaporo
'won (currency)' 'water polo'
Wben lwl is a post-consonantal glide, as in (78), then it often~ iu/. 46 With donor lkw gw kw gwl, post-consonantallwl may~ iwl and we find the monomoraic glides /1-...>vV gwVI in the pronunciation of some contemporary speakers (79). Other speakers will pronounce the words in (79) as ikuootarii/, etc. Exceptions to the patterns in (78--79) include those in (80). Here, post--consonantallwl ~ 0. (78)
Eng. guava Eng. switch Eng. quark
(79)
Eng. quarterly Eng. quintet Sp. Guatemala
kwootarii k\-viNteQto gwatemara
'quarterly (magazine etc.)' 'quintet' 'Guatemala'
(80)
Eng. equal Eng. squash Eng. sweater Eng. Qantas Eng. bilingual
ikooru sukaQsu seetaa kaNtasu bairiNgaru
'equivalence, =' 'squash (drink & sport)' 'sweater' 'Qantas' 'bilingual'
guaba SttiQCi kuooku
'guava' 'switch' 'quark'
'Ihe voiceless labio-velar lA·\ I, strictly speaking a fricative and found for example in Scottish English, typically~ /how/ before a substituted /a/ (81a) and~ /ho/ before /i e/ (8lbc). I know of no examples of this phone occurring before substituted /o/ or /u/. In view of the lack of contact with English speakers who have this phone, this substitution must be categorized as a dictionary tradition
46. See,howeve1; the examples in (139-141) illustrating alternative, though more restricted, pattern-s of substitution.
Chapter 3. Phonology 93
whose origin is probably late 19th century prescriptive Received Pronunciation lhwl (Mugglestone 2003: 186···189). All gaimigo containing an element whose source is Eng. white (and some gairaigo dictionaries list up to 30 such entries) are pronounced as in (81a), regardless of the fact that English varieties with IMI or lhwl in this word are now in the minority. (81)
a. b.
'white' Eng. white howaito 'whistle' hoiQsuru Eng. whistle Eng. whale watching hoeeruuoQCiN gu 'whale watching'
The palatal approximant Lil ~ lyi before substituted or epenthetic ia o u/ (82a-c), but~ 0 before substituted /i/ (82d) . .Befixe substituted /e/, three different patterns are found. The first, and by far the most common, is Ul ~Iii (82e). For some contemporary speakers, and generally confined to names, UI may~ ly/ (82f). Other speakers \·\'ill pronounce words such as (82f) \·\'ith initial/if, i.e. as I ieena/, following the (82e) pathway. Finally, Ul may also~ 0 (82g), primarily in Russian loans. This last pattern is the product of a dictionary tradition. (82)
a. b. c. d. e. f. g.
Eng. yard Ger./odel Kor. yukhoe -4·§1 Eng. yeast Eng. yellow card Ger. lena Ru. EKaTepnH6ypr jjekatieriinburkl
yaado yooderu yuoke iisuto ierookaado yeena ekateriNburuku
'yard' 'yoder 'raw minced beef with egg' 'yeast' 'yellow card' 'Jena' 'Yekaterinburg'
\Alhen Ul is a post ..consonantal glide (e.g.lbjl), or 'Nith palatalized consonants (e.g.lbil), 47 then we witness the same patterning as (82). This is illustrated in (83). Here, though, before the problematic substituted ie/, Ul either~ Iii (83e) or 0 (83f). The glide /ye/ is not found, even in innovative contemporary speech. (83)
a. b. c. d. e. f.
Ru. PH3aHb lriazanil ryazani fyoodoru Ru.
'Ryazan' 'Fyodor (Dostoevsky)' 'puma' '(Alexander) Pushkin' 'sovief 'Belarus'
The close frontness of UI means that it causes the preceding donor consonant to behave as if it were followed by iii. Thus, as we have already seen in (17, 19 ..·20,
47·
Por palatal, rather than palatalized, consonants see (25, 44,71-72, 74-75).
94
Loanwords in Japanese
29, 41, 49) above, ltj tJ tsj tsil ~ /C/, I dj dj zj eli (ji zil ~ /j/ and IOj sj Oi sil ~ iSi. This is illustrated in (84). Before substituted /a u o/, ldj dil t)'lJically do not~ /j/, but ~idyl (85). (84)
Ru. IIynm lputiinl Ru. Me.n;Be)J;eB lmiedvied!efl Ru. Ko3HHIWB lkoziincefl Eng. Matthews
(85)
Eng. duet Eng. duty free
puuciN medobeejefu koojiNcefu masuuzu
dyueQto dyuu tiifurii
'(Vladimir) Putin' '(Dmitry) Medvedev' '(Grigori) Kozintsev' '(Stanley) Matthews'
'duet' 'duty free, tax free'
The labio-palatal approximant Ill I displays the same substitution pattern as the vowellyl to be described in (97 -99). Examples include kyuijiinu 'cuisine' ~Fr. cuisine and cuintrii ~Fr. Tuileries.
Finally, donor geminate consonants typically~ !QC/, as illustrated in (86). The substitution of geminates is more complex in borrm\'ings from Korean, for which see Sugawara (2006: 22···24). (86)
Ar. lmrek:rel
It. espresso
meQka esupureQso
'Mecca' 'espresso'
The foregoing consonant substitutions are summarized in 'Htble 3.2. As adaptations may ditTer according to the .f()llowing substituted vowel, five diilerent gai· raigo moras are shown for each of the donor phones listed in the two vertical shaded columns. Donor lm n I) I may also ~ IN! in certain environments. The symbol '>' indicates a diachronic transition from an earlier conservative to a later contemporary mora. Multiple entries in a cell indicate multiple substitution pathways. Table 3.3 is a reverse reading of Table 3.2. Gairaigo consonant phonemes are shown on the top ro\-\~ their possible sources below. Common source phones are in bold.
3.2.1.2
ViJwels
Substitution of donor vowels by the five Japanese vowel phonemes /i e a o ui is schematized in Figure 3.4 and illustrated in (87 -91 ). 'Ihese examples are typical synchronic pathways- exceptions exist in many cases. Unmapped I+ Y y 0 oe <)I are
roblematic: see (93-100, 113, 116-120). Unr donor words.
1 any
(87)
lil--+/i!
III --+/if (88)
lei --+ /e/ lei-+ /e/
Fr. kilogramme Eng. gin Eng. sense Fr. filet
kirogu jiN SeNSU hire
able 3.2. Summary of consonant substitutions IPI
pa
pi
pu
pe
po
lxl
lbl
ba
bi
bu
be
bo
111'1
It I
ta
Ci>ti
cu>tu
te
to
lhl
ldl
da
de>di ji>di
zu>du
de
do
Ihi
Ill
ja
ji
ju
je
jo
ldzl
lkl
ka
ki
ku
ke
ko
ltfl
lgl
ga
gi
gu
ge
go
ld3l
lql
ka
ki
ku
ke
ko
It~ I
IGI
ga
gi
gu
ge
go
lt~l
cu
lei
Its I
1?1
a
u
e
0
ldo.l
1~1
ba
bi
bu
be
bo
lpfl
1£1
ha>fa
hi>fi
fu
he>fe
ho>fo
lml
Ivi
ba>va wa>va
bi>vi ui>vi wi>vi si
bu>vu U>VU
be>ve ue>ve we>ve se
bo>vo UO>VO wo>vo
In I
so
IJll
1111
ji
zu
I!JI
lsi lzl
si ji
su zu su su
lrl IRI se>Se
ju
je
jo
Ill lA. I
181
su
IJI
Sa
131
ja
ji
17J
ja
ji
ju
je
jo
Ill lwl
191
ha
hi
fu
he
ho
IMI
lxl
ha
hi ki
fu
ho ko
ga
gi
gu
he ke se ge
Iii
ka
go
141
1~1
IYI
su
96
Loanwords in Japanese
Table3.3. Sources of gairaigo consonant phonemes
p
b
t
d
k
g
c
l
j
f
v
s
z
s
h
m
n
T
w
p pf
b v
t
d
k
g
ts
tJ
J
n
l
t
Jl
n
y
ts
z
X
s 9
w v
X
X
B.
M.
.1..
I)
G
t!O
('j
~
X
IJ dz d
h f
m
t
s 9
z
I)
f h
v
q
~ d
tj d3 z
~
J
~
Jl H
r
h
.1..
3
c
dz
X h
~
y
X
pf
M.
N
m
-q
~
y
t~
y
j
r::~~::::J ------i Front Close
Nea~ front
Central
' ----;-!- - - !
Near b~ck Back ,!
Close mid Mid
~ ------------ ,
r-- ---Open mid
l
/o/
l
L__________j
Near open Open
Figure 3.4. Summary of vowel substitutions
The donor vowels (89) lal and (90) lvl require further explanation. These are typically the General American (GA) and Received Pronunciation (RP) realizations of the same English phoneme. Thus, lot is GA [lat], RP [lvt] . Since the former -4 /a/ and the latter -4 /o/, confusion is not surprising. To an extent, this confusion can serve as evidence for a UK or US origin of a given loanword. Since (89) soccer is rarely used in the UK (instead football), it appears as expected to be a borrowing from GA lal. On the other hand, the footballing term (90) offside
Chapter 3. Phonology
would appear to be a borrowing from RP lvl. Similarly, volleyball, a sport introduced from the US, ~ baree(booru) from GA lal, while the tennis term volley, a sport introduced from the UK,~ boree from RP l11l. There are even cases of the same \·\'ord being borrowed from both RP and GA, an example being koNfareNsu ~ kaNfareNsu, both 'conference: Undue weight should nevertheless not be attached to these two vowel pathways as a means of distinguishing British English from American English borrowings. As mentioned in §2.4.4, the greater number of RP .. speaking English foreign advisors resident in Japan in the late 19th century when dictionary traditions \'\'ere formulated has meant that it is the adaptation /o/ of the RP realization lvl that is the more frequent. There are even a number of loans which, although their US provenance is almost certain, nevertheless~ /o/.mther than /a/: e.g. hoQtodoQgu 'hotdog: sukoQciteepu 'sticky tape' ('Scotch Tape' in the US, 'Sellotape' in the UK), wisukoNsiN ''Wisconsin'. (89)
lrel-t /a/ lal-t /a/ IBI-t /a/ IAI-t /a/ lal-t /a/48
(90)
lol-t /o/ 1:>1--t io! lvl - t !of
(91)
lul-t /u/ lwl-t/u/ lul-t/u/
Eng. map Ger.Kapsel Cant. wan 4tan 1 ~;e:. Eng. cut Eng. soccer Fr. chapeau Fr. culotte Eng. offside
saQpo kyuroQto ofusaido
Ger. Urethan Kor. taegi1kki Ell:·.; 7] Eng. looks
maQpu kapuseru waN taN kaQto saQkaa
'map' 'capsule' 'wonton dumpling' 'cut' 'football, soccer'
'hat' 'culottes' 'offside (in football)' uretaN teguQki rttQkusu
'urethane' 'name of Korean flag' '(good) looks'
The near-open front vowel (89) lrel frequently triggers a preceding dose front glide /y/, although only when the preceding phone is a velar plosive. In (92) are listed just a few examples of many. Although this is likely a diction .. ary tradition, Lovins (1973: 72) notes that, experimentally, 'the frontness of Ire I is normally perceived by speakers of Japanese' and that this frontness 'is most readily interpreted as palatalization of the preceding consonant. .. particularly... when it is a back consonant'. There are a small number of examples where no glide is found (e.g. ga_jeQto ~ Eng. gadget, katarogu ~Eng. catalogue) and a few gairaigo having variants with and without a glide (e.g. kamem .., kyamem ~ Eng. camera).
48.
The sixth donor vowel in this group, ]a], only occurs long, as ]3:]. See (115).
97
98
Loanwords in Japanese
(92)
lsrel-+ /gya/
Eng. caddie Eng. cabin .Eng. career Eng. guara(nteed fee) Eng. gallery Eng. gamble
'caddy' 'cabin' 'career' 'performance fee' ~;yara 'gallery' gyararii gyaNburu 'gambling'
kyadii kyabiN !..')·aria(a)
I now move on to the problematic vowels, unmapped in Figure 3.4. The close central unrounded vowelli:l is found only in a few donor words from Russian, as shown in (93). It is typically substituted as /uii, but occasionally also as ii/. (93)
'(Andrei) Gromyko' Ru. rpoMblKO lgrom4kol guromuiko '(Vasily) Smyslov' Ru. CMbiClTOB lsmislofl sumisurofu komima- korima 'Kolyma' Ru. KorrhiMa lkolimal
Dictionary traditions govern the substitution of the mid front rounded vowels Ire ol. In borrowings from German, lcrl regularly~ iel (94), but in borrm·\'ings from French lcrl ~/of (9Sa), while lrejl ~ /iiyu/ (95b) or /uiyu/ (95c). On the other hand, lol in borrowings from French typically~ /u/ (96a), while in borrowings from other donor languages 101 may~ /o/ (96b) or~ /e/ (96c). For further analysis and discussion of the asymmetry in the substitution of lol and Ire I from French and German, see Dohlus (2005a). Note also the long vm\'el substitutions of 10:1 in (101) below. Ger. Koln Ger. Rontgen
keruN reNtogeN
'Cologne' 'x-ray'
(94)
Ire I-+ /e/
(95)
a. b. c.
'hors deeuvre' Fr. hors··d'reuvre oodoburu lrel-+ /o/ 'millefeuille' lcejl ---+ /iiyu/ Fr. millefeuille miru.fiiyu Fr. trompe·i'oeil toroNpttruiyu 'trompe··loeil' lcrjl ---+ iuiyu/
(96)
a. b. c.
lol-+ iui Fr. un·deux··trois aNdutorowa '1··2·3' toromuso 'Tromso' lol-+ /o/ Nor. n·oms@ 'raw beef with egg' lol-+ /e/ Kor. yuktwe ~§I yuQke
As shown in (97), the dose front rounded IYI and the near-close near-front rounded IYI both ~ lyui, thus combining close frontness (ly/) 'Nith roundness (lu/). As illustrated in (98), however, donor ltY ty tsY tsy ~fy tfyl ~ /cu/, lsY sy fY fyl ~ isu/ and lzY zy 3Y 3YI ~ ijui, all with no /y/-glide (cf. 17, 19, 29, 49, 84). As shown in (99), donor ldY dyl ~ idyu/, not !jui (cf. 85). (97)
lnYI-+ /nyu/ Ger. Niirnberg nyuruNberuku 'Nuremberg' '(Albert) Camus' kamyu lmyl ---+ /myu/ Fr. Camus 'Utrecht' Du. Utrecht yutorehito I#YI-+/yu/
Chapter 3. Phonology
Ger. Tiibingen Ger. Zurich Fr. Saussure Ger. Sd1Utz Ger. Suskind Fr. Jules
(98)
ltYI --+leu! ltsYI --+leu/ Isri-+ /sui lfY I --+ I sui lzYI --+ /ju/ I3YI --+ tjul
(99)
'(Alexander) Dumas' dyuma Idyl-+ /dyu/ Fr. Dumas IdYl --+ /dyu/ Ger. Dusseldmf dyu.Qserudorufu 'Dii.sseldorf
cu. biNgeN cuuriQhi sosuuru SUQCU juusukiNto juuru
'Til bingen' 'Zurich' '(Ferdinand de) Saussure' '(Heinrich) Schi.itl '(Patrick) Si.iskind' 'Jules (Verne)'
A dictionary tradition operates in cases where German IYI is spelt (y), found mainly in words German has itself borrowed from Greek. As shown in (100), here IYI ~ /i/ probably; as noted by Dohlus (2005b:42), because of earlier Japanese loans from Dutch and Portuguese where orthographic (y) ~ /i/. (100)
Ger. Hysterie Ger. Synthese
hisuterii jiNteeze
'hysteria' 'philosophical synthesis'
The substitution of donor long vowels largely follows the patterns outlined above, i.e. IV :I typically ~ /VV i, with the quality of iVI based on the adaptation pathways illustrated in (87-100). 49 In (101) I restrict exemplification to the most commonly found donor long vowels and treat English free (tense) monophthongs as long, as they are in RP (Wells 1982: 117 -124). Pathways f(n 10:1 vary according to the dictionary tradition of the donor language. (101)
li:l--+ Iii/ ly:l --+ iyuu/ le:l --+lee! la:l--+ /aai la:l--+ /aa/ 1::.:1--+ ioo! lo:l --+ /oo/ lu:l --+ /uu/ l0:l--+ /eel 10:1--+ /ee/ 10:1--+ /uu/ 10:1--+ /oo/
Eng. scene Fr. luge Ger. 'lhema Ger. Gaze Eng. spa Eng. straw
SiiN ryuuju teem a gaaze supaa sutoroo Ger.Neumse noirooze Eng. blues buruusu Ger. Goethe geete Hu. Gyor jeeru Fr. Chartreuse sarutonmzu Sw. Norrkoping noosoopiNgu
,
'scene 'luge' 'topic' 'dressing, gauze' 'spa' '(drinking) straw' 'neurosis' 'blues (music)' '(Johann von) Goethe' 'Gyor' 'Chartreuse' 'Norrkoping'
49· For the phenomenon known as 'pre-nasal shottening, where a long vowe!IV:I /VI before a tantosyllabic lnl, see (105-108).
--7
short
99
100
Loanwords in Japanese
Besides the lowering of ldi:l ~/dee/ noted in (20d), li:l ~ /eei in a few other borrowings from English, though generally restricted to cases where the spelling is (ey). Since this substitution is sporadic, a dictionary tradition is likely not at \·\'ork here. The examples listed in (102) are best seen as spelling loans. (102)
li:l
~!eel
volleyball hockey curry money medley Silicon Valley Harley- Davidson
bareebooru hoQkee karee manee medoree sirikoNbaree haareedabiQdosoN
'volleyball' 'hockey' 'curry' 'money' 'medley' 'Silicon Valley' 'Harley-Davidson'
For English words exhibiting the cross .. dialectal variationlju: "' u:l (e.g. due, nude, lewd, etc.), it is the variant with the palatal glide that has typically been borrowed. Thus nyuusu, not *nuusu, 'news: ryuuto, not *ruuto, 'lute'. ,....._,
The substitution of donor diphthongs may or may not follow the pathways of their constituent vowels. A selection of the most common donor diphthongs which do are shm·\'n in (103), of those which do not in (104). (103)
larl ~ /ai/ laul ~/au/ l:)rl ~/oil
Eng. line Eng. pouch Eng. toilet paper
(104)
lerl ~!eel ldul ~/oo/ I:)YI ~ ioi/ lceyl ~/oi/
Eng. lace Eng. show Ger. Neurose Du. spuit
raiN pauCi toireQtopeepaa
reesu soo noirooze supoito
'line' 'pouch' 'toilet paper'
'lace' 'show' 'neurosis' 'pipette'
As pointed out by Ichikawa (1929: ll, 1930: 180), Lovins (1973: 81 ..·82) and Kubozono (1995a, 1999b: so ... 51), there is a strong tendency for donor lerl ~ ie! (105) and laul ~/a/ (106) before tautosyllabic In I. This phenomenon is knm'\'n as pre-nasal shortening, its motivation lying in the general avoidance of superheavy syllables (14, p. 75). The phenomenon also applies, though less strongly, to donor ldul, as well as to long vowels in general (107). However, as noted by Kubozono (2001a: 66, 2002a: 86), shortening does not occur with donor larl, as the example in (103) above shows. There is evidence that the tendency towards pre-nasal shortening appears to be weakening. Some gaimigo which exhibit the phenom .. enon are obsolete or becoming so. AisukuriN 'ice cream' f- Eng. ice cream, for example, has been largely replaced by aisukuriimu, although it may still refer to
Chapter 3. Phonology
a traditional type of ice cream found in Okinawa and elsewhere. Other gaimigo have unshortened doublets: e.g. (106) g-.trauNdo and fauNdedoN. Importantly, there are many examples of gairaigo "'"here pre-nasal shortening seems never to have operated, especially when donor lnl is word-final. Some of these are shown in (108). Eng. range Eng. stainless (steel)
(lOS)
leml-7 /eN/
(106)
launl -7 /aN/
Eng. ground Eng. foundation
(107)
launl-7/oN/ liml-7 /iN/
Eng. ionely Eng. kerosene
(108)
leml-7 /eeN/ launl -7 /auN/ launl-7/ooN/ liml -7 !iiN/ lu:ni-7/UUN/
Eng. lane Eng. count Eng. loan Eng. jeans Eng. spoon
'stove, hob' 'stainless steer
reNji suteNresu guraNdo faNdeesoN
roNrii kerasiN reeN kattNtO rooN jiiNZU supuuN
'sports ground' 'foundation'
'lonely' 'kerosene' 'lane' 'count' 'loan' 'jeans' 'spoon'
,.._, Nasalization in donor vowels typically -7 IN!: i.e. IVI-7 /VN/. The most com·· mon patterns are shown in (109). There are, however, three major deviations from the vowel pathways shown in (87-100): l~rench lsi -7 /aNi, not */eN/; 50 l~rench lli";j -7 /aNi, not "/oN/; Portuguese lnul -7 /aN/, not */auN/. These are illustrated in (110). If pre--nasal shortening (105···108) is extended to nasalization, then the motivation for the last of these three deviations is clear.
(109)
lol-7 /oN/ 1:>1-7 /oN/ lal-7 /aN/
(110)
lel-7 /aN/ lcel-7 iaN/ li3ul-7 /aN/
Por. Rondonia Fr. marro11 Fr. vacarzces Fr. dessin Fr. \'erdun Por. Silo Paoio
roNdonia maroN bakaNSU
'Rondonia 'Spanish chestnut' 'vacation'
deqsaN berudaN ~· verudaN saNpauro
'drawing, design' 'Verdun:' 'Sao Paolo'
50. 'Ihe few Polish names containing lei do not, unlike donor Prench lei, deviate from the pathway in (88): e.g. ceNsutolwva 'C:l{Stochowa' and vaweNsa '(Lech) Wal~!f.
101
102
Loanwords in Japanese
Rhoticization of vowels in English is complex and shows considerable variation across dialects: see Wells (1.982: 213···245) for a thoroughgoing description. However strong the influence of American varieties of English may have been on Japanese since the early 20th century and despite the pm·\'erful role of orthography in adaptation patterns (§3.2), it is non-rhotic RP, regarded as the standard in England, on which substitutions have been overwhelmingly based. 5 1 As noted in §2.4.4, this is in all probability due to the prestige accorded the RP pronunciation of British 52 foreign advisors, who comprised the majority of English-speaking advisors when dictionary traditions were beginning to take shape in the late 19th century.53 In all the examples of rhotic vowel substitution to follow (111-ll.S), two sets of donor phones are shm·\'n. To the left of a swung dash ~- the GA articulation is given, to the right the RP. Both donor pronunciations are those cited by Wells (1982: 11.8-124). Adaptation patterns often diiier according to whether a rhotic vowel appears in syllable-final position (111···113) or elsewhere (114···115). Patterns may be further subdivided according to donor rhotic vowel type. Donor syllable-final rho tic vowels may be split into three types: (A) those where RP shows a schwa---final diphthong IV<"ll, or an underly:ing54 schwa---final triphthong IVV<"ll; (B) those where RP shows a long VO'Nel articulation; and (C) the GA rhotic schwa monophthong la-1. Non-syllable-final rho tic vm-\'els may be divided into two types: (A') those where RP shows a schwa-final diphthong IV<"ll, or underlying schwafinal triphthong IVV<"ll; and (B') those where RP sh(Y1NS a long vowel articulation. The situation is summarized in Table 3.4.
The few rare exceptions include ooru ~-··· oar (0 1906) and moorusu ~-··· Morse (Code) (':-1869), but these are early borrowings, as their .first written attestations show. Other famous Morses may be moosu, especially later ones such as Colin Dexter's .Inspector .A1orse detective novels, well-known in Japan. What appear to be further exceptions are typically borrowings from other languages: e.g. yooguruto 'yoghurt' from Ger. Joghurt, karuteru 'cartel' from Ger. Kartel or porunogurafii (> poruno) 'pornography' from Ger. Pornographie or Fr. pornographie.
_51.
What proportion of British nationals was Scottish or Irish is unknown. It is unlikely they would have spoken RP.
52.
.53· Mutsukawa (2009: 57-76) takes an opposing view. His synchronic theoretical analysis of 'English ir/ realizations' takes auditory GA forms as the input. 54· Such underlying triphthongs may be subject to smoothing and have various realizations among RP speakers, including the triphthongic realization shown here (Wells I 982: 238-242).
Chapter 3. Phonology
Table 3.4. Adaptation of English rhotic vowels Donor syllable-final lm ~ uJI ~ /ua/ or luaal lu ~ rJI ... ~ /i(y)al or /i(y)aa/ lt:l ~ e;)l ~leal or leaa! Iau;,.-- ml;)l----~ /awaa/ lara.~ ar;)l ~ lai(y)a/ or /aiyaa/
A (111)
lal'" a:l ~ /aa/ l:.l ":1 ~ /oa/
B (112)
Type
Elsewhere
A'
Iu -- I;)l ----~ /i.a/ or /iaal
lu1 ~ u;)l
(114)
N
101'" ::.:1 ~loa/
c
IJ. -- ;)I ~ /aa/ or /a/
H' (115)
(113)
~
lu(w)aa/
l~;1 ~ €JI ~leal or /eaa/ Iau..~~ atJJI----) /awaa/ lai~-~ aiJI ~ /aia/
lal-- u:l ~ /aa/ l.:u ~ ::.:1 ----~ loo/ lm ~ ::.:1 ~ loo/
11 ~ a: I ----~ /aa/
Type A diphthongs~ /Va(a)/, as shown in (llla-c), while Type A triphthongs pattern as illustrated in (lllde). Here, the lVI and IVVI segments of the RP schwa .. final diphthongs and triphthongs ~ /V/ or /VV/ according to the adaptation patterns in (87 -91). The exception is (111d), where donor lui ~ /w/ rather than expected lui. Meanwhile, the finall<>l segment of the diphthongs and triphthongs ~/a/ or /aa/. Although there appears to be an increasing tendency towards the shorter /a/ substitution, more research is required to clarify a confused situa-tion. The palatal glide /y/ may be inserted after iii in (111be). A iy/-glide is occasionally found after /e/ (111c), but this is uncommon. For example, although some speakers use heya 'hair; hea is more widespread. (111)
a.
overture tour lu ~- I<'ll ~!ia/ clear career ~ /iaai ~ /iya/ gear ~ /iyaa/ shear ~lea/ shareware le.T- e<'ll ~ /eaa/ sports1-vear lauJ.-- au<> I ~/awaa/ shower sapphire larJ.-- m<>l ~ /aia/ wire ~ /aiyai ~ /aiyaa/ hire lm Nual
~/ua/
~/uaa/
b.
c. d. e.
oobaacua cuaa
'(musical) overture' 'tour'
kuria kyariaa giya Siyaa
'clear' 'career' 'gear, transmission' 'shear'
'shareware' seawea supoocuweaa 'sportswear' sawaa
'(bathroom) shower'
safaia waiya haiyaa
'sapphire' 'wire' 'hire'
Type B vowels pattern as in (112). Donor RP 1:):1 does not~ /oo/, as in (101). Instead, the adapted form /oa/ appears to reflect a donor "I:)<> I f(nm which, though found in neither modern RP nor GA, did occur (for the lo1 ~- :)ll vowel only) in 19th century RP (Trudgill2010: 106).
103
104
Loanwords in Japanese (112)
lal- a:l l=>l =>:1 lol -:>:I
~faa! ~loa!
par and/or
~/oa/
door
paa aNdoa do a
'(golf) par' 'and/or' 'door'
The syllable-final Type C rhotic vowel, 1<>-1 in GA but 1;;,1 in RP, typically ~ /aa/ (ll3a). Also not infrequent, especially with more recent borrowings, are cases where doublets in both /aa/ and /a/ exist (l13b). (113)
a.
Ia.- al ~ /aa/
batter anchor calendar
b.
I;,.~,
Jl
~
'batter' 'anchor' 'calendar'
baQtaa aNkaa kareNdaa
iaa ,_a! computer koNpyuutaa '" koNpyuuta 'computer' printer puriNtaa " puriNta 'printer' ele~·ator erebeetaa - erebeeta 'lift, elevator'
1ype A' diphthongs ~ /Va(a)/, as shmvn in (114a-c), while Type A triphthongs pattern as in (1 14de). TI1e lVI and IVVI segments of the RP schwa-final diphthongs and triphthongs ~/VI or /VV/ according to the adaptation patterns in (87 -91). Again, the exception is donor lui, which~ /w/ rather than expected /u/ (114d). The final IJI segment of the diphthongs and triphthongs ~ Ia/ or /aa/, the latter /aa/ substitution being more common. A labial glide /w/ may be inserted after /u/ in (ll4a). No insertion of the palatal glide /y/ is found after /i/ in (114be). Contrast this with the glide insertion found in syllable-final Type A vowels in (lllbe). (114)
a.
jul
b.
lu ,_ I;JI
U;)l
~
/uaa/ Stewart, Stuart sucuaato ----) /uwaa/ steward sucuwaado piene Sears
piasu siaazu
'pierced ear, etc: 'Sears (Tower)'
/ea/ ---) /eaa!
Caims public affairs
keaNzu
'Cairns'
~/awaai
Coward
kawaado
'(Noel) Coward'
----)Jam,
iron
aiaN
'(golfing) iron'
~ ~
c.
d. e.
leJ- e;)l jau<>
~-
au;JI
laia." ai;JI
'Stewalt, Stuart' 'steward'
/iai /iaa/
~
'
-
I
paburiQkuafeaazu 'public affairs'
Type B' vowels f()Jlow the long vowel patterning outlined in (101), while ~ /aa/. Thus:
11 ~- a:l
(115)
lru '" a:l l::u - :):1 lo.l - :>:I 11 - 3:1
~faa!
mark
~ /oo/
fork
~ /oo/ ~ /aa/
pork network
maaku fooku pooku neQtowaaku
'mark' 'f()rk' 'pork' 'network'
To return to the summary in Table 3.4. Since Type B'
11
~
.:nl and Type C
1<>- ~,;;,I are in complementary distribution, the former may be reassigned as Type C'
This affords us the broader synthesis of English rhotic vowel adaptations shown in Table 3.5.
Chapter .3. Phonology
Table 3.5. Adaptation of English rhotic vowels: A broader synthesis
Type A-A'
Anypo~tion
Notes
luJ-u~l ~!ual
/ua/ may be /uaa/ anywhere lual may be /uwaa! when not donor syllable-final /ia/ may be /iaa/ anywhere /(a)ia/ may be /(a)iya/ or /(a)iyaa! when donor syllable--final /ea/ may be /eaa/when donor syllable-final /oo/ is loaf when donor syllable-final
lu- ~~1 ····~ /iai lel- e;)l ~leaf
lau;J.. ·- au~l····~ /awaa!
lai;J.- aJ;)I ~ /aia/ B-B'
l
C - C'
13- - ;)I ~ /aa/ 11 ·- 3:1 ~ /aa/
/aa/ maybe /a/when donor syllable-final
The majority of German dialects are, like English RP, non-rhotic. What is written (r) non-prevocalically in words such as Hamburg, Saarland, Berlin and Dortmund is generally articulated [-e] (Kohler 1977: 177). Nevertheless, Japanese patterns of substitution typically fc)llow a dictionary tradition based on less com-· mon rhotic dialects. Thus, haNburuku, zaaruraNto, beruriN and dorutomuNto, as well as (17a, 35, 44, 73, 97, 99, 121, 193) and the vast majority of relevant examples listed in Table 2.9. The exception appears to be with schwa, written (er) and generally articulated [-e] (opus cit.). This typically -4/aa/. Examples include (32b, 70, 128, 129), as well as countless names such as wagunaa '(Richard) Wagner' and a.dena.uaa '(Konrad) Adenauer'.
,-._, I turn finally to the remaining cases of donor schwa Ia I. These are governed by a range of dictionary traditions. Spelling loans are not uncommon. Donor German lal typically -4 /e/ (116), while French lal -4 /u/ (117). In English, the substitution of schwa lal often follows that of the donor spelling (118), although exceptions abound, as illustrated in (119). As shown in (120), a dictionary tradition applies to English loans ending in lfanl, when written (tion) or (sion). (116)
Ger. These Ger. Geliinde
(117)
Fr. Bretagne burutaanyu Fr. reportage ruporutaaju
(118)
a.
b. c.
d.
lal-+ lei lal-+ ial lal-+ /o/ lal-+ /u/
teeze gereNde
'thesis' 'ski slope, piste'
garden .final police symposium
'Brittany' 'documentary' gaadeN fainaru porisu siNpojiumu
'garden 'final' 'police' 'symposium'
105
106
Loanwords in Japanese
(119)
a. b. c. d.
(120)
lf<1nl-7 isoN!
3.2.2
1<>1 1<>1 1<11 1<>1
(e) --7 /u/ (ai) --7/e/ (o) --7 /u/ (e) --7 Iii
open Britain iesson kitchen
auction option junction impression
oopuN buriteN reQSUN kiQCiN
'open' 'Britain' 'lesson' 'kitchen'
ookusoN opu8oN jaNkusoN iNpureQSON
'auction' 'option' 'junction' 'impression~
Epenthesis
Vowel epenthesis is examined in §3.2.2.1, mora obstruent epenthesis in §3.2.2.2. 3.2.2.1
Vowel epenthesis
As noted in §3.1, the Japanese syllable must be either open or have a coda in /N/ or /Q/. The dosed syllables and consonant clusters found frequently in donor words are adapted to Japanese phonotactics by means of vowel epenthesis. Which vowel will be inserted is determined by the value of the donor consonant. The epenthetic vowel is, most commonly, one of the two dose vowels /u/ or Iii, both of which are conspicuous in Japanese in that they undergo regu·· lar devoicing in certain environments (Maekawa & Kikuchi 2005; Tsujimura 2007: 23-27; Vance 2008a: 206-214). By far the more frequently encountered of these two epenthetic vowels is /u/, the Japanese vowel most subject to weak· ening and deletion, as well as being the shortest phonetically (Sagisaka & Tohkura 1984). Below I illustrate vowel epenthesis for the donor phones in Table 3.2. Not included are lc q G '(X t~ • ,{ w 1•\ ql. I am unaware of any donor words in 'Nhich these phones occur in an environment where they would attract an epenthetic vowel, i.e. either directly before another consonant or word-finally. Non-prevocalic III found in many English varieties is treated as rhotacism of the preceding vowel: see (111···115) for detail. TI1ose donor consonants- the majority- which take epenthetic /u/ are: (121)
IPI---7 /pu/ lbl ---7 /bu/ lkl ---7 /ku/ lgl---7/gu./
Eng. ,1,>roup Eng. pub Eng. screen Eng. green 1~1···7 /bu! Sp. Pablo ltl ····~ /fu/ Eng. phrase Ivi ---7 /bu " vu/ Eng. eve I91 ····) /sui Eng. thriller
gunmpu
pabu sukurii..c" guriiN
paburo fureezu
ibu" ivu suriraa
'group' 'pub' 'screen' '(golfing) green' 'Pablo (Picasso)' 'phrase' 'Christmas Eve' 'thriller'
Chapter .3. Phonology ICJI ----> !zu! lsi~ /stt/ lzl ~/zu/ 1n ----~ /sui
131 ~/ju/ IIrl~ /ru/ lhl----~ /fu/
itsl ~/cu/ ldzl~izu/ IP~ ~/pu! lml ~/mu/ l11l ~ /NFft/ lrl ~inti IRI----~
Ill~
/ru/ /ru/
Ul ~lyu!
Eng. rhythm Eng. status Eng. cruise Eng. flash Fr. beige Fr. encore Fa1: loobrruedi:ne:p:dl Eng. pants Eng. kids Ger. Schwarzkopf Eng.fJm Eng. aqualung Ru. 6opru; lborsc'l Ger.Hormon Eng. pool Fr. •Warseille
rizumu suteetasu kuruuzu furaQsu beeju a.lll"kooru afumadiinejaado paNcu kiQZU suwarucukoQpu firumu akuaraNgu boru8ici horumoN puuru maruseeyu
'rhythm' 'status' 'cruise' 'flash' 'beige' 'encore' '(Mahmud) Ahmadinejad' 'underpants, trousers' 'kids' '(Elizabeth) Schwarzkopf 'film' 'aqualung' 'borsht' 'hormone' 'pool' 'Marseille'
Donor In I only attracts an epenthetic /u/ under very restricted circumstances: word--finally in borrowings from French (60---62). Although donor lml and IIJI always attract an epenthetic /u/ when donor word-final, as illustrated in (121), they do not necessarily do so before a consonant. Recall that lml 4 /N/ before a bilabial or labiodental (63---64), while IIJI typically 4 IN! before a velar or alveolar plosive (69). With donor word-finallkl, early borrowings often show an epenthetic /i/. A selection of these is shown in (122). Some of these do have a later doublet with epenthetic lui, but typically from a different polyseme: e.g. jcrQku 'jack, knave (in cards)', sutoraiku 'strike (in baseball)', sutiQku 'hockey stick, drum-stick', etc. (122)
[0+1811] [ 0 .1833] [0 1857] [+1864] [0 +1867] [0 +1874] [0 1874] [0 1879] [ 0 1883] [0 1899]
Du. biik Du.kalk Du.dek Eng. cake Eng. steak Eng. stick Du.pek Eng. strike Eng. jack Eng. brake
buri(Q)ki karuki deQki keeki suteeki suteQki peNki sutoraiki jaQki bureeki
'tin(plate )' 'bleaching powder' 'deck' 'cake' 'steak:' 'walking stick, cane' 'paint' '(labour) strike' '(car) jack' 'brake'
'Ihe donor cluster lksl often 4 /kis/ or /kiS/, with an epenthetic /i/, as illustrated in (123). Although there exist doublets with both epenthetic /i/ and /u/, in most cases one variant is overwhelmingly more frequent. For the three exam-ples shown in (124), tekisuto outgoogles tekusuto 118:1, takusii outgoogles taki.5ii
107
108 Loanwords in Japanese
640:1, and ekusukuramedoNmaaku outgoogles ekisukuramedoNmaaku 22L1:1. 55 Examples with an epenthetic lui only no epenthetic Iii variant are generally restricted to cases where the donor lksl cluster is ""·ord·· or morpheme--final. This is illustrated in (125) and (126), respectively. Even here, though, (125) seQkusu [ 0 +1885] has an earlier attested doublet, seQkisu [+1877]. Adaptation patterns dearly have been, and still are, volatile. (123)
Du. extract Eng. Texas Eng. expert Eng. exciting
(124)
Eng. text
ekisu tekisasu ekisupaato ekisaitiNgu
'extract, essence' 'Texas' 'expert' 'exciting'
'texf 'ibid: 'taxi' Eng. taxi 'ibid: Eng. exclamation mark ekisukurameeso Nmaaku ekusukurameesoNmaaku tekisuto tekusuto takisii takuSii
(125)
Eng. box Eng. sex Eng. fax
(126)
Eng. spokesman Eng. boxer Eng. sexy
boQkusu seQkusu faQkusu
'ibid:
'box' 'sex' 'fax>
supookusumaN bokusaa sekusii
'spokesman' 'boxer' 'sexy'
With donor lx y hi, we find dictionary traditions whereby the epenthetic vowel depends on the preceding donor vowel. vVhen this is an open front vowel, the epenthetic vowel is /a/, as shown in (127); with other front vowels, the epenthetic vowel is /if, as illustrated in (128). When the preceding donor vowel is a close back vowel, or where there is no preceding vowel, the epenthetic vowel is /u/, as shown in (129); with other back vowels, the epenthetic vowel is typically /o/, as illustrated in (130). It appears that in borrowings from Russian and Polish the epenthetic vowel following lxl is always /u/, regardless of the quality of the preceding donor vowel. The few examples are shown in (131). (127)
55·
lxl-7/ha/ Ger. Bach baQha '(Johann Sebastian) Bach' Ger. Eisenach aizenaha 'Eisenach' lnl-7/ha/ Ar.lfatahl fataha 'Fatah'
When the words were searched as a katakana'exact wording or phrase' on 5 July, 2010.
Chapter 3. Phonology
(128)
lxl--+ /hi/
Ger. Liechtenstein Du. Maastricht Du . .Anderiedlt Ger. BUchner
(129)
lxl--+ /fu/
'(.Eduard) Buchner' Ger. Buchner bufunaa Ger. Brudt buntQftt '(Maxo) Bruch' Du. Groningen furooniNgeN 'Groningen'
IYI --+ /fu/
rihiteNsutaiN maasutorihito aNderurehito byuuhinaa
'Liechtenstein' 'Maastricht' 'Anderlecht' '(Georg) Buchner'
(130)
IYI--+ /ho/ Du. van Gogh baNgoQho '(Vincent) van Gogh' lxl --+ /ho/ Gae. loch roQho 'loch'
(131)
lxl--+ /fu/
Ru. Ka3axcTaH lkazaxstanl kazafusutaN 'Kazakhstan' Ru. AntaTOBa laxmatoval afumaatova '(Anna) Akhmatova' Pol. Lech refu 'Lech (Wah;sa)'
With retroflex, alveolo-palatal and palatal donor phones, the epenthetic vowel is Iii, as illustrated in (132). This is also the case for palatalized consonants, of which I offer only a selection in (133). Donor IJ11 from French attracts an epen·· thetic lui, for which see (71). (132)
(133)
IJI --+Iii! 1~1 --+I Sit 1~1--+ /ji/ lal --+/Cit llfl --+/hi! IJll --+/nil lmil~imii
lt(Oil ~!Cit lril ~ /ri/ llil ~ /ri/
Hu. Nagy lnvJI Ru. DymKHH lpuskiinl Ru. 5pe)KHeB lbrieiniefl Pol. L6di Ger. Zurich Pol. Gdansk
naji puusikiN burejinefu UQci cuuriQhi gudanisuku
'Nagy (Imre)' '(Alexander) Pushkin' '(Leonid) Brezhnev' 'L6di' 'Zurich' 'Gdansk'
perumi Ru. IIIoCTaKOBH'I lsostakoviiCil sosutakoviQci caari Ru. I~apb lcaril Ru. ApxaHreJJI>CK larxangie~skl aruhaNgerisuku RtL IlepMb lpennil
'Perm' '(Dmitri) Shostakovich' 'czar' :-\.rchangel'
It is unclear why the postalveolar affricates ltf d3l should also attract an epenthetic Iii, as shown in (134), despite the postalveolar fricatives If 31 attracting an epenthetic /u/ (121). Lovins (1973: 1.22-126) notes Carr's (1951) 'principal of consonantal resonance' and the 'greater nativeness' of lsi! over lsu/. Some early borrowings with If 31 did take a 'more native' epenthetic /i/, although many of these have more recent doublets with an epenthetic lu/. TI1is is illustrated in (1.35). (134)
ltfl--+ rc.il ld:jl --+ /jil
(135)
1.0
.Eng. inch Eng. page
iNci peeji
'inch' 'page'
····~ lsi/ ·-
burasi ~ buraQsu isu! Eng. brush saqsi saQsu Eng. sasl1 Eng. hashisl1!Ger. Haschisch hasiSi hasi(Q)su potaaji ~ potaaju I:~ I --7 /ji/ ~ /ju/ Fr. potage N
N
'brush' 'window frame' 'hashish' 'soup, potage'
109
no Loanwords in Japanese
With It dl, the epenthetic vowel is typically ioi, as shovm in (136). In the conservative phonemic system, It d! have affricate/fricative allophones before the two close vowels /u i/: see (11.). An epenthetic /o/ preserves /t d/ as plosives. Th.e motivation for /o/, as opposed to an equally preserving /a/ or /e/, is, however, unclear. (136)
ltl --+/to/ ldl-+ /do/
Eng. straight Eng. android
sutoreeto aNdoroido
'straight' 'android'
Despite /tu/ and /du/ being permissible moras in contemporary pronunciations (18-19), there has been no shift to an epenthetic lui, i.e. there are no contemporary pronunciations such as (136) *suturetu or *aNduroidu. The /u/ of a contemporary /tu/ or /du/ mora is thus never epenthetic, but always a substitU·· tion of an original donor vmvel. vVhile there are a number of earlier loans where the epenthetic vowel was iu/, here It dl \'\'ere not preserved as the plosives it d/, but --+ /c z/. The examples in ( 137) do not include any donor words which may have been borro'~Ned from an English plural (e.g. guQzu 'goodS, ruucu 'roots, ori .. gins: niizu 'needs, demands: supiricu '(alcoholic) spiritS, etc.) or from items that are infrequently employed in the singular (e.g. buucu 'boo( biizu 'bead: naQcu 'nu( etc.). Th.ese may be~ Its dzl rather than~ It dl. Some of the gairaigo in (137) do have a later doublet with epenthetic /o/, but typically from a different polyseme: e.g. siito 'sheet (of paper): bakeQto '(crane) bucket, bowling term'. (137)
In/a] [•1864] [•1873] [•1884] [ 0 1885] [ 0 1899] [ 0 1905] [ 0 1910] [ 0 1911] [ 0 1920] [ 0 1925] [ 0 1930]
Ru. KaM•raTKa lkami::iatkal Eng. cutlet Eng. shirt Eng. sheet Fr./.Eng. omelette Eng. bucket Eng. fruit Eng. suitcase Eng. Christmas tree Eng. drawers Eng. settlement Eng. suit
kamucacuka kacurecu sacu siicu omurecu bakecu furuucu suucukeesu kurisumasucurii zuroosu securumeNto suucu
'Kamchatka' 'cutlet' 'shirt' '(bed) sheet' 'omelette' 'bucket' 'fruit' 'suitcase' 'Christmas tree' '(women's) drawers' 'settlement (work)' 'suit'
Although an epenthetic /o/ is found in most borrowings with tautosyllabic ltwVI, as in (138), this may not be the case with tautosyllabic ltwel where, as shown in ( 139), doublets in /cue/ are found. As illustrated in (140), tautosyllabic ltwi twi:l typically -'t /cui cuii/. Compare (78). (138)
ltwVI-+ /towV/
Eng. twirler Fr. toile
towaraa towaru
'(baton) twirler 'toile'
Chapter 3. Phonology n1
(139)
itwel ~/cue~· toe/
Eng. twelve Eng. twenty
(140)
itwrl ~/cui!
twin
Itwi:l --7 Icuiil
Eng. Eng. Eng. Eng.
Twitter tweed tweeter
toerubu -- cuerubu toeNtii ·-· cueNtii
cuiN cuiQtaa cuiido cuiitaa
'twelve' 'twenty'
'twin' 'Twitter' 'tweed' 'tweeter'
Finally, an epenthetic /o/ occurs after any consonant, not only jtj, in a French donor string jCwaj. These~ /Cowa/ or /Coal: (141)
lmval ~ /rowa/ lfvval ~ /fowa/ lmwal ~ /rnoa/ lpwal ~/poa/
croissant foiegras moire Poitiers
kurowaQsaN fowagura moare poatie
'croissant' , 'foie gras 'moire' 'Poitiers'
........, Vve have already seen that there are a small number of older gairaigo with an unexpected epenthetic Iii after donor lkl (122), or /u/ after donor It dj (137). The earliest attestations for such loans date predominantly from the late 19th and early 20th centuries. If one goes back even further in time, the description of epenthetic vowels outlined in this section becomes less robust. With Iberian borrowings (§2.2), a pattern of anticipatory replication can be found. This is shown in (142). Here, the epenthetic vowel is the same as that found in the subsequent mora, boxed in the examples. According to Sawada (1985: 84---87), this pattern is particularly prevalent when the final consonant in a donor cluster is a liquid, as is the case with all the examples in (142). Epenthesis based on anticipatory replication is not, however, the dominant pattern. With other Iberian borrowings, epenthetic /u/ is more frequently found: see Tables 2.3 and 2.4. (142)
a. b. c. d. e. f.
Por. credo Por. cristao Por. profeta Por. padre Por. sacramento Lat. ecclesia
ke~do kir{I]sitaN pmiQiheeta batet{§'N sakar@jmeNto eker~sia
'credo' 'early Japanese Catholic' 'prophet' , 'priest, father 'sacrament' 'church'
Anticipatory replication can also found in borrowings from the Dutch period: (143)
a. b. c.
Du. glas Du. trap 'stairs' Du. strychnine
gar~su tar@jQpu sutorikin[!]ine
'(pane of) glass' 'airstairs, gangplank' 'strychnine'
n2
Loanwords in Japanese
The examples in (142ef) and (l·Bc), rewritten in (144), may also be viewed as perseverative replication, whose occurrence is less common. Here, the epenthetic vowel is the same as that found in the previous mora. (144)
Por. sacramento Lat. ecciesia Du. strychnine
s~arameNto ~eresia
su tor[!]kiniine
'sacrament' 'church' 'strychnine'
By the period of Dutch borrowing, replication, be it anticipatory or perseverative, had become rare and patterns of epenthesis were gradually giving way to those found today. Vestiges of replication are apparent only in the epenthetic vowels found after lxl in modern borrm.·vings from Dutch, German, Arabic and Gaelic (l27-l30), as well as in the epenthetic vowels of a few older English and French borrowings: (145)
[0 [0
1874] 1907] [0 1909]
Eng. salad Eng. truck Fr. croquette
sar~a tor~Qko > toro ko~Qke
'salad' 'handcar' 'croquette'
For early literature on the historical replication processes outlined in (142--145), sometimes referred to as 'vowel harmony' or 'assimilation', see Ichikawa (1930: l83) or Doi (1933: 1L1-15).
A1ora obstruent epenthesis When the mora obstruent /Q/ occurs in canonical /V_C/ position, the /QC/ component is essentially a long, or geminate, consonant exhibiting 'suspension of articulatory movement' (Komatsu 1980:566) or 'prolongation of consonantal articulation' (Shibatani 1990: l67 168). This 'suspension' or 'prolongation' of the succeeding obstruent creates an extra mora. In traditional Japanese linguistics the mora obstruent !Q! is known as sokuon ilt ifj·. The motivation behind /Q/-epenthesis is complex. As the donor language is of particular importance, a cross--linguistic analysis in the same vein as those pro-vided in previous sections of this chapter would be inapposite. I will therefore restrict initial discussion to English gaimigo and treat borrowings from other languages thereafter. What follows below is intended as an overview of a phenomenon still the object of much ongoing research. 56
3.2.2.2
;6. Much of the analysis which follows owes a debt to previous research by Ohye (1967), Ohso (1971), Lovins (1973), Quackenbush (1977, 1989), Aoki (1981), Watanabe & Hirato (1985), Koo & Homma (1989), NINJAL (1990), Hirata (1990), Ono (1991), Hirozane (1992), Takagi & Mann (1994),Kanai (1995), Kawagoe (1995), Ura (1995), Kitahara (1997), Katayama
Chapter 3. Phonology 113
Epenthesis of the mora obstruent /Q/ occurs in three major environments, WORD-FINAL, STRESSED MEDIAL and SYLLABIC L, and in one minor environment, woRD-FINAL CLUSTER. In none of these environments is /Q/ generated exceptionlessly and, as in earlier sections, what I describe here is the dominant synchronic patterning. In many cases, previous descriptions of these environments have defined the consonants before which /Q/-epenthesis occurs as 'obstruents' (plosives, fricatives and affricates). This is erroneous. Although never occurring before sonorants, the patterning we find is too refined for such a broad term as obstruent. In order to make subsequent analysis clearer, Table 3.6 summarizes the presence or absence of !QI -epenthesis in each of the three major environments, listed in the leftmost column, before each of the English obstruents, listed in the shaded rows. A tick .r indicates that I Q/ is inserted with overwhelming frequency, a cross X that /Q/-epenthesis rarely if ever occurs, and a boxed tick 0 or cross lEI indicates a strong tendency towards or away from /Q/-epenthesis. An empty box D indicates an approximately even split, while a blank cell indicates that no examples exist. English donor Its dzl are perceived as affricates by Japanese speakers and are included in my analysis. One English obstruent, the voiceless glottal fricative Ihi, does not occur in any of the relevant environments and so is not treated further. Table3.6. I Q/ -epenthesis in loanwords from English across three major environments
ts
tf
./
./
./
~
./
~
p
t
k
f
9
s
f
WORD-FINAL
./
./
./
)1.
~
~
~
IBl IBl
)1.
STRESSED MEDIAL
~
~
SYLLABICL
./
IBl d
./
./
b
g
v
WORD-FINAL
IBl
~
~
)1.
STRESSED MEDIAL
)1.
IBl
)1.
)1.
)1.
SYLLABIC L
)1.
)1.
IBl D
)1.
)1.
)1.
lS
./
IBl
z
3
)1. )1.
./
dz
d3
./
./
)1.
IBl )1.
Across all environments, /Q/-epenthesis occurs only after a checked vowel, never after a free vowel, a rhotic vowel, or a diphthong. Checked vowels in English are 'subject to the phonotactic constraint that they do not occur in a stressed monosyllable with no final consonant' (Wells 1982: 119) and are almost
(1998:69-149), Yamane & Tanaka (2000), Yamane (2001), Maruta (2001), Kawagoe & Arai (2002), Preston & Yamagata (2004), Rice (2006), Kubozono (2007), Kawahara (2008), Otaka (2008: 123-155) and Hirayama (2008).
l14 Loanwords in Japanese
identical to, and sometimes also termed, the lax vowels. They are II reA ul and what is lei in RP and lei in GA. A sixth vowel lvl, checked in RP, corresponds to GA lui, where it is free. Like the substitution of rhotic vowels (111 115), it is RP articulations on which /Q/-epenthesis is based and thus it may be triggered after lv ~ al. Moreover, /Q/-epenthesis occurs only when a checked vowel is substituted by a short vowel in Japanese, a substitution which occurs with overwhelming fi-equency. The Japanese syllable in which /Q/-epenthesis occurs is thus a heavy one, (C)(G)VQ. There are, however, both dictionary loans (e.g. imeeji ~ image, etc.) and spelling loans (e.g. derikeeto ~delicate, etc.) where a checked vowel is substituted by a long vowel. Here, /Q/-epenthesis never occurs and f(xms containing a superheavy syllable, (C)(G)VVQ, such as *imeeQji or *derikeeQto, are never found. 'Ihis is in line with the claim that superheavy syllables are typically eschewed in Japanese (Kubozono 1989, 1995a, 1999b).
TI1e first major environment in which epenthesis of the mora obstruent !Q! typically occurs in borrowings from English, the WORD-FINAL environment, is where a donor word ends in a plosive or affricate preceded by a checked vowel. In other words, donor words containing the string: (146)
IVC#I, where lVI word boundary
=
checked monophthong, ICI == plosive or affricate, # ==
Such a string typically~ /V1QCV/, where IV) is an epenthetic vowel (§3.2.2.1). This is illustrated in (147) and (148). (147)
(148)
tip
CiQpu daieQto kuriQku gaQCU sukeQci
IVp#l IVt#l IVk#l IVts#l IVtf#l
click guts sketch
IVd#l IVg#l IVdz#l IVd3#l
thoroughbred smog odds judge
diet
'tip, gratuity' 'diet' '(mouse) click' 'gut~, courage' 'sketch:'
sarabureQdo sumoQgll OQZU jaQji
'thoroughbred' ' smog
)
'odds' 'judge'
English donor words with WORD--FINAL lbl are an exception to the pattern in (146). Here, /Q/ is rarely triggered. Some of the very few examples where it is include 11JOQbU 'mob: SUriOQbU 'snob' and sukyaQbu 'scab (worker)'. for donor words in finalldl, /Q/ is generated with only some exceptions. While /Q/ is also typically triggered f(H donor words in finallgl, there are a number of examples
Chapter .3. Phonology 115
where it is not, including bumgu 'blog' (although bumQgU is occasionally found), gyagu '(comedy) gag' and anarogu 'analogue: The reason for this d > g > b ranking among the voiced plosives, where donor words with finalldl are the most likely, and those with finallbl the least likely, to trigger /Q/-epenthesis is unclear. Koo & Homma (1989: 130) believe it may be due to the closure duration of /b/ being longer than that of ld! or /g/, and the vowel preceding lb/ thus being shorter. Katayama (1998: 127 128) speculates on a link with intervocalic spirantization of the voiced plosives: only /dl 'does not weaken'Y The possibility also exists that we are dealing with dictionary traditions. In a conservative pronunciation, donor ICI in (146) may be devoiced during adaptation, so that the examples in (148) become sarabureQto, sumoQku, etc. In a survey of 40 informants, Quackenbush ( 1989: 5-7) found that the youngest (all aged 12) and oldest (aged 35-45) age cohorts were most likely to devoice the final consonant and pronounce words such as beQdo 'bed' and burudoQgu 'bulldog' as beQto and burudoQku. Although she did not survey donor word-finallbl and her oldest age cohort is comparatively young, the curve elicited by her results closely mirrors that proposed by Downes (1984: 191) to describe the relationship between vernacular usage and age. In other words, contemporary forms such as beQdo, used more frequently by Quackenbush's 20---25 year--old age cohort, appear to be more prestigious than the vernacular conservative forms such as beQto used by the youngest and oldest age cohorts. 58 When a donor English word ends in a fricative in WORD--FINAL environment, the mora obstruent /Q/ is, for the most part, not generated. This is illustrated in (149). With donor lfl there is at least one common counterexample with epenthetic /Q/: s:utaoju '(member of) staff'+-- staff. Donor lfl is the exception. Here, !Q/ is regularly found, as shown in (150). (149)
(150)
off love
IVf#l IVv#l IVEl#l IVs#l IVz#l
jazz
ofu rabu sumisu basu jazu
IVJ#I IVf#l
cash stylish
kyaQsu sutairiQsu
Smith bus
'off' 'love' 'Smith' 'bus'
'jazi 'cash' 'stylish'
57· Historically, the equivalent voiceless ranking t > k >pis found when describing the extent (least to most common) of word-internal allophonic voicing among the Old Japanese plosive tenues (Unger 2004). 58. Quackenbush (opus cit.) also fow1d that overall devoicing occurred more commonly in conversation-style than in word-list or reading-style elicitations.
n6 Loanwords in Japanese
To the author's knowledge, the only attempt to statistically quantify the level of /Q/-epenthesis in WORD·FINAL environment has been that of Maruta (2001:73---75), who used the gairaigo dictionary SSDH (1991) f(H his database (n = 972), but excluded proper nouns, mora-dipped forms (§3.4) and compoundclipped forms (§4.3.1.1). Since his analysis is not carried out by donor ICI but by substituted /C/, I have converted and in some cases amalgamated his figures. 59 These are summarized in Table 3.7. Table 3.7. Percentage of/Q/-epenthesis by English donor obstruent in environment (after Marut.a 2001) English word--final d.ono.r phone % /Q/-epenthesis
wORD-FINAL
tf
3id3
p
t
k
f
d.
g z/6idz ts
b
f
s/0 v
100
100
99
99
98
91
71
56
23
13
1 0
50
33
Maruta's analysis confirms the ranking d > g > b amongst the voiced plosives. The unexpectedly low incidence of 33% shown for donor ltsl is problematic. Maruta includes words such as .~acu 'shirt' in his analysis of substituted /c/, but this donor word actually has final ltl preceded by a rhotic vowel. If other such words were excised from Mantta's corpus, the figure f()r English donor ltsl would likely rise dzl is contribconsiderably. The incidence of 50% for donor word--final lz ~ uted to solely by ldzl. I know of no examples where !Qi is triggered before donor English word-finallol or lzl.
a--
'Ihe second major environment in \·vhich /Q/-epenthesis occurs is STRESSED MEDIAL. Here, epenthesis is triggered when an intervocalic voiceless obstruent is preceded by a stressed checked vowel. In other words, donor words containing the string: (151)
IV1CV2 1. where ICI =voiceless obstruent, IV 11"'stressed checked monophthong, IV 2J "' any vowel
Such a string typically~ /V 1QCV /· This is illustrated below: (152)
happy batter saccharin
haQpii baQtaa saQkariN
'happy' 'batter' 'saccharin'
59· Thus, donor lsi and lSI are amalgamated, since both 7 /s/ or IS! (28-29,40-41). Similarly, l:il and ld:il both 7/j/ (42, 52), while lzl, lol and ldzl all 7 /z/ or /j/ (28-29, 40-41, 50).
Chapter 3. Phonology 117
IVltsV21 IVttfV21 IV1ElV21 IV1sVzl IVJV21
sprftzer pitcher nothing message admission
supuriQca pioeaa naQsiNgu meQseeji adomiQsoN
'spritzer' 'pitcher' 'nothing' 'message' '(university) admission'
Exceptions in STRESSED MEDIAL environment are far more prevalent than in woRD--FINAL and can easily be found for most of the voiceless obstruents listed in (152). Some are listed in (153). Lack of iQ/-epenthesis is especially apparent before lfl in donor words ending with {tion). Despite the number of exceptions, /of--epenthesis unquestionably occurs more often than not in STRESSED MEDIAL environment. (153)
IVlpV21 IVltV21 IVlkV21 IVltfV21 IVlsV21 IV tlVI 2
copy bUtter lfquor natural
classical ambition
kopii bataa rikaa nacuraru kurasikaru aNbiSoN
'(photo )copy' 'butter' 'liquor' 'natural' 'classical' 'ambition:'
In STRESSED MEDIAL environment, /Q/ is generated only very rarely before lfl and typically not generated at all before voiced obstruents. Some examples where !Q! is generated are shown in (154). These are largely restricted to the co -occurrence of a following -ing or -er morpheme. (154)
IV1tv2l IVldV21 IVlgV21 IV1d3V2I
bUffer heading I' s~ugger
dredger
baQfaa heQdiNgu suraQgaa doreQjaa
'buffer (in computing)' 'header (in football)' 'slugger (in baseball)' 'dredger'
Ohso (1971: 33, cited in Lovins 1973: 119-120) has pointed out that since the string /QCVQC/ is generally avoided in Japanese, 60 potential conflicts are resolved by an environment ranking whereby WORD--FINAL dominates STRESSED MEDIAL. This may be illustrated in the f{)llowing examples: (155)
Gothic ketchup ticket
gosiQku kecaQpu cikeQto
'Gothic (art, font)' 'ketchup' 'ticket'
6o. Exceptions are most common w~here this string straddles a morpheme boundary (e.g. noQtoQta 'was in accordance with', fliQpaQta 'tugged', (198) pauiQkuQta 'panicked: seQkeQkyuu 'erythrocyte').
n8 Loanwords in Japanese
Here, forms such as *goQsiku or *goQsiQku, with STRESSED MEDIAL /Q/epenthesis, are not found. A few obsolete gairaigo are exceptions: e.g. [+>1919] koroQkeQto 'croquette'.
I turn now to the last of the three main environments, SYLLABIC L. Although environment is similar to STRESSED MEDIAL, /Q/-epenthesis here is triggered before a more restricted number of voiceless obstruents, typically when IP k f s tfl are followed by word--final syllabic I and preceded by a stressed checked vowel. In other words, donor words containing the string: SYLLABIC L
n
(156)
IVq#l, where ICI = IP kfstfl, lVI "'stressed checked monophthong,# =word boundary
Such a string typically-4/V 1QCV2 ru/, where IV/ is an epentheticvowel. Examples are shown in (157), although those containing donor ltfl are few and far between. (1.57)
jVpl#l IVkl#l IVfl#l jVsl#l IVtJl#l
apple tackle w4ff7e whistle lvfftchell
aQpuru taQkuru waQfuru hoiQsuru miQc.eru
'apple' 'tackle' 'waft1e' 'whistle' '(David) Mitchell'
Cases of /Q/-epenthesis before donor ltl and lfl in SYLLABIC L environment are scarce (158a, 159a), examples such as those in (158b, 159b) without any /Q/epenthesis being the norm. Borrowings containing the English donor string IVg! #j are rare and occur both with (160a) and without (160b) /Q/--epenthesis, as well as doublets in both (160c). /Q/-epenthesis in SYLLABIC L environment appears never to occur before other obstruents. (lS8)
IVtJ#I
a. throttle b. shrtttle
(159)
IVfl#l
a. busl1ei b. initial
(160)
IVg)#l
a. j>lggle b. t6ggie
stnlggl.e
suroQtoru satoru buQseru inisaru
'throttle, gas' 'shuttle' 'bushel' 'initials'
jaQguru to guru sutoraQguru
sutoraguru
'juggle' 'toggle' 'struggle'
Although it is tempting to extend the SYLLABIC L environment to cover English word-final syllabic II) I. here /Q/ -epenthesis is erratic, examples fewer, and variation in the epenthetic vowel found after donor ICI much in evidence:
Chapter 3. Phonology 119 (161)
IVt~:;t#l IVt~:;t#l IVk~:;t#l
IVtf~:;t#l IVsr~#l IVf~:;t#l
cotton mitten chfcken kitcl1en lf.sson muffin
koQtON mit oN cikiN kiQciN reQSUN mafiN
'cotton' 'mitten' 'chicken' 'kitchen' 'lesson' 'muffin:' .---.._,
The minor environment which triggers epenthetic !Q! is WORD-FINAL CLUSTER, where a donor word ends in a plosive--fricative cluster preceded by a checked vowel. In other words, donor words containing the string: (162)
IVCpCF#I, where ICPI =plosive, ICFI =fi·icative, lVI =checked monophthong,#= word boundary
Such a string typically~ /V 1QCV2 CV/, ·where vowels: (163)
IV/ and IV/ are both epenthetic
'box' boQkurm IVks#l box jeepoQpusu 'Japanese pop music' IVps#l J-pop(s) IVpe#l depth (gauge) deQpusu (geeji) 'depth gauge'
The English plosive-fricative clusters ltj ts d3 dzl are perceived as affricates and thus trigger /Q/-epenthesis according to WORD-FINAL environment (146). /Q/--epenthesis does not occur to any extent in WORD-FINAL CLUSTER environment when the consonant cluster is composed of two ti-icatives, two plosives or with the reverse order fricative-plosive. ,-.__,
In cases where a loanword may be perceived as consisting of 'two prosodically independent. .. morphemes' between \·\'hich 'an interval. .. is inserted' (Ura 1995: 180), then '#' in (146) WORD-FINAL, (156) SYLLABIC L and (162) WORDFINA.L CLUSTER environments may be redefined as 'morpheme boundary'. Here, a considerable amount of I Q/--epenthesis may be found: (164)
hatchback haQcibaQku 'hatchback' upgrade aQpugureedo 'upgrade' SYLLABIC L duffle coat daQfurukooto 'duffle coat' applemint aQpurumiNto 'applemint' WORD--FINAL CLUSTER tax-free taQkusufurii 'duty--ti·ee, ta..x-free' Oxbridge oQkusuburiQji 'Oxbridge' wonD-FINAL
120
Loanwords in Japanese
Table 3.8. /Q/-epenthesis in loanwords from English: summary by environment and manner of articulation voiceless voiceless voiceless voiced voiced voiced all all plosives fricatives affricates plosives fricatives affricates voiceless voiced WORD-FINAL
STRESSED !viEDIAL
0
[B)
D
/i
D
/i
/i
/i
Iall
!phones
D
D
10
0
11
!D
[B) /i - - - - - - -0 - - - 0- - - - - - - - - - - - - - /i' · · @.............J .............J D SYU.ABIC I.
AI. I. ENVJRONM.ENTS
0
D
[B)
)I
D
D
Table 3.8 groups and averages the contents of Table 3.6 by manner of articula · tion (plosives, fricatives and aftricates, both voiceless and voiced). The same symbols are employed. While the averages are not weighted in any way and thus take no account of the higher .frequency of certain phones in certain environments, they do serve as a rough index on which may be based some general observations on /Q/-epenthesis in English gairaigo. These are: a.
b.
d.
iQ/-epenthesis is most likely to occur in WORD-FINAL environment, where it is almost always triggered before affricates (voiceless./, voiced./) and voiceless plosives (./).It has been argued that 'English word--final voiceless plosives after lax [checked] vowels have, to a Japanese ear, an auditory impression which is closer to geminate consonants than to single ones' (Takagi & Mann 1994: 345). Nevertheless, /Q/--epenthesis also occurs frequently in WORD··FINAL environ·· ment before voiced /d gi (148, Table 3.7). Across all environments and articulations, /Q/-epenthesis occurs with appreciably greater frequency before voiceless phones (0) than before voiced (lEI). Across all environments it is triggered most commonly before voiceless afhi·· cate (./) and voiceless plosive (0) articulations. Th.e only examples of signif.icant /Q/-epenthesis before voiced phones are in WORD-FINAL environment before aftricates (./) and plosives (D). Epenthesis never occurs before voiced fricatives (x). Voiced affricates display the most variation by environment, with /Qi--epen· thesis occurring almost always(./) in WORD-FINAL environment, but hardly ever (X) in STRESSED MEDIAL and SYLLABIC L environments. Voiceless fricatives display the least internal consistency as an articulatory class. In SYLLABIC I. environment, we find a strong tendency to-
Chapter .3. Phonology
/Q/-epenthesis (1&:1), with /Q/ found to any degree only before lfl. In this environment, Ohye (1967: 114-115) has claimed /Q/ is not generated with the other three English donor voiceless fricatives Is f 91 because they are 'attracted to the previous mora'. For a discussion on a possible link which I discount--- between /Q/--epenthesis and double letters in English spelling (i.e. (tt) (pp), etc.), see Arai & Kawagoe (1996, cited in Preston & Yamagata 2004) and Shirai (1999). In other environments, /Q/epenthesis is extremely sporadic and awaits further research.
Examples of /Q/-epenthesis in loanwords from languages other than English are fewer, the data meagre and syntheses problematic. I offer here a brief description of I Q/-epenthesis in loanwords from French, German, Dutch, Russian, Korean and Portuguese. There are no examples of the phenomenon in Chinese gairaigo. Although the term 'checked vowel' is generally applied only to English, redefining lVI in (146, 156, 162) and IV 11 in (151) as 'short' or 'lax' allows us to examine the four environments cross-linguistically. Apart from the distinction between lei and le:l made by some speakers, vowel length in French is non--phonemic. One condition under which allophonic lengthening occurs is when lo 0 ul appear in closed stressed syllables (Tranel1987:49). In French, word-final syllables containing these vowels are always stressed. Thus, they are always long in WORD-FINAL environment and examples of /Q/-epenthesis in gairaigo where the donor vowel is lo 0 al are consequently rarely fi-mnd. Also rarely found are examples where the donor vowel is lui. This vowel appears to resist /Q/-epenthesis and instead undergoes compensatory lengthening (Shinohara 1997: 81). 1bus, eNkuruuto 'en croftte' f- I~r. en crotUe and guadoruupu f- Fr. Guadeloupe. With other French donor vowels, WORD-FINAl. /Q/--epenthesis is possible and some examples are shown in (165). Nevertheless, compensatory vowel length-ening may be found with these vowels too (Arai 1993). Examples include aNkeeto 'survey' f- Fr. enquete and esukabeeS-.t 'escabeche' f- Fr. escabeche. As illustrated in (166), /Q/-epenthesis in WORD-FINAL environment generally does not occur before lsi and ln. Nor does it occur before voiced obstruents. Here, again, we typically find compensatory vowel lengthening (Miyashita 1993), as illustrated in (167). (165)
lak#l laf#l lek#l let#l lik#l lip# I
Fr. Balzac Fr. Camn di\che Fr. Quebec Fr. baguette Fr. esthetique Fr. Louis-Philippe
baruzaQku karaNdaQsu kebeQku bageQto esutetiQku ruifiriQpu
'(Honore de) Bal1..ac' 'Caran d'Ache' 'Quebec' 'baguette' 'beauty care, beauty salon' '(King) Louis- Philippe'
121
122
Loanwords in Japanese
(166)
(167)
lis#l les#l laf#l
Fr. l'v1atisse
lib#l lev#l la3#l
Fr. .Antibes Fr. Geneve Fr. reportage
Fr.Metz Fr. Piaf
matisu mesu piafu
'(Henri) .Matisse' 'Metz' '(Edith) Piaf'
aNtiibu juneebu juneevu ruporutaaju N
~Antibes'
'Geneva' 'documentary>
Since stress in French typically falls on the final syllable, candidate gairaigo for STRESSED MEDIAL /Q/·epenthesis are scarce. However, stress falls on the penulti·· mate syllable when a 'Nord ends in syllabic IT!I and here we do find loans with /Q/ epenthesis: e.g. riQtoru 'litre' f- Fr. litre and its compounds. On the other hand, meetoru 'metre' f- metre and its compounds exhibit compensatory vowel lengthening instead. Yet other gairaigo have medial /Q/ despite syllable··final stress: e.g. deQsaN 'drawing, design' f- Fr. dessin, eQ_{eru 'Eitlel (Tower)' f- Fr. (Tour) Eiffel. Although Miyashita (1993: 218-219) believes examples like these are spelling loans (Figure 3.3) influenced by double letters, there exist cases which are clearly not influenced by donor spelling. These include saQpo 'hat' f- Fr. chapeau; see the summary in Arai (1993: 80-85). French gairaigo appear not to undergo WORDFINAL CLUSTER /Q/-epenthesis (168), while the lack of any examples means comment on SYLLABIC L /Q/-epenthesis is not possible. (168)
IVks#l IVks#l
Fr.A.ix Fr. .Asterix
ekusu asuterikusu
'Aix' 'Asterix'
Borrowings from German and Dutch show similar patterns of I Qf.·epenthesis to English. Here, /Q/ is typically generated in both WORD-FINAL (169ab) and STRESSED MEDIAL (169cd) environments and, though examples are scarcer, in \\TORD-FINAL CLUSTER (169ef) and SYLLABIC L environments (169gh) also. (169)
a. b. c. d. e. f.
g. h.
IVx#l IVk#l IV.tV I ' 2 IVltV21 IVps#l IVks#l IVkl#l IVsl#l
Ger. 1vfach Du.dek Ger. Hutte 'hut, cabin, lodge' Du. letter 'letter' Ger. Schnapps Ger. Felix Du. nikkel Du. Brussel
maQha deQki hyuQte reQteru sunaQpusu feriQkusu niQkelu buryuQseru
'J\tlach (speed)' 'deck' 'mountain(eering) hut' 'label' 'schnapps' 'Felix (Mendels.<;ohn)' 'nickel' 'Brussels'
With Russian loans, /Q/ -epenthesis is rare. It is typically absent in both WORDFINAL and STRESSED MEDIAL environments. When the final syllable is stressed, a similar phenomenon to that found in French may be witnessed in WORD·H.NAL environment, that of compensatory VO'Nellengthening: e.g. kuraaku 'kulak' fRu. KyJiaK !kulak!. The only environment where /Q/-epenthesis may occur in
Chapter 3. Phonology
Russian appears to be with the suffix -OBWI -lovlii':il, 'son of: and its variants (170a). Epenthesis may occur in personal names from other Slavic languages before this su11ix, 'Nhere it has various pronunciations, spellings and translitemtions (170b). There are, however, many examples where epenthesis does not occur and many family names and patronymics have doublets with and without /Qi. (I 70)
a.
Ru. PoC1porroBH•I jrostropoviiol rosutoropomriQci. '(Mstislav) Rostropovich'
b.
Serb. MilosevicMHnornem·d'l
mirosebiQci
'(Slobodan) MHoseviC'
With loans hom Korean, according to Sugawara (2006: 24-26) and from whom the examples below are taken, /Q/ is typically not triggered in WORD-FINAL environment when the final consonant is IPI or lkl (171ab), although there are sporadic cases where it is (17lcd). It never occurs when the final consonant is ltiHere, ltl ~ !Qi (17le). Since Korean is not a stress--timed language, conditions for the STRESSED MEDIAL environment do not exist. Nevertheless, Suga-
a. b. c. d. e.
IVp#l IVk#l IVp#l IVk#l IVt#l
]oil. Kor.pap }.l tll.., Kor. pak Kor. ssampap ·'t:rlfl tl'l ..., Kor. ttok .£d::. Kor. tolsof '2. 'El
papu palm saNpaQpu toku -- toQku tOrtlSOQ
'rice, meal' 'Park (Chung-hee)' 'wrapped rice' 'sweet rice cake' 'stone pot'
'Ihere are at least two examples \·vhere I Q! -epenthesis occurs in Iberian borrowings. All are from Portuguese, occur in STRESSED MEDIAL environment before donor Ipi, and are the oldest examples of /Q/-epenthesis in the gairaigo stratum: (172)
3.2.3
[•1608] [ 0 16551
Por. capa Por. copo
kaQpa koQpu
'raincoaf 'suit of cups'
Deletion
Deletion is by far the least significant of Japanese adaptation strategies since, for the most part, it is the product of an auditory source (.Figure 3.3). As I have remarked already (§3.2), such borrowing has been infrequent. Smith (2006) offers a theory-based account of deletion and cites a score or so of examples, to which I append below not only further cases, but also dates of first written attestation. As Smith notes, most gaimigo which have undergone deletion pos-· sess doublets \·vhich have not. Where extant, these 'full forms: which have an orthographic source and have undergone substitution and epenthesis, are given below each example of deletion. Many examples cited below, both of deleted and full f(Hms, are obsolete or obsolescent: e.g. (1.73) raNbiki, (1.74) bisuteki, (175)
oorumito, yaarudo.
123
124 Loanwords in Japanese
Smith divides her examples into five categories; I make some readjustments and offer four. The first of these is deletion of an initial unstressed donor vowel, where examples are fe\,,. (173). The second is donor consonant duster simplifica·· tion (174). Where cluster simplification involves deletion of a \·\'ord-final consonant, it is listed under category three. This category, the largest, contains examples of donor word··final consonant deletion and is illustrated in (175). (173)
Por. alambique Eng. American
raNbiki merikeN 61 amerikaN
'alembic' 'America(n)' 'American'
(174)
Eng. Hepburn
heboN heQpubaaN
'(James) Hepburn' 62 (1815···1911) '(Audrey) Hepburn' (1929-1993)
Eng. glycerine
risuriN guriseriN
'glycerine' 'ibid:
[0 [0
Eng. beefsteak
bisuteki63 biifusuteeki
1)eefsteak' 'ibid:
[0 1874] [+1872]
[0 [0
1686] 1854] [ 0 1912]
Eng. white shirt waiSacu 'shirt' howaitosacu 'ibid: (175)
1912] 1900]
Eng. alright
oorai ooraioorai 64 ooruraito
'alrighf, backing a vehicle [•1864] 'alright alright' [0 1877] 'alright, 0 K' [0 1914]
Eng. lemonade
ramune remoneedo
'lemonade' 'ibid;'
[0 [0
Eng. mast
roosu roosuto
'sirloin' 'roast meat'
[0
Eng. dozen
daasu
'dozen'
[0
'handkerchief' 'ibid: 'ibid;'
[•1882] [•1898] [+1871]
Eng. handkerchief haNkeci haNkaCi haNkeruCiifu
61.
[0 [0
1872] 1873]
1870] 1905]
.1871] ,.1864] 1877]
Now only found in compounds: e.g. the hybrid merikeNko 'flour:
American missionary and compiler of the first Japanese-English dictionary (§2.4.4), after whom the hebonshiki romanization system is named (§5.1).
62.
63. Bifuteki [0 1883] is likely not another example of donor consonant cluster simplification, but borrowed from Fr. bifteck 'steak: itself from Eng. beefsteak. 64. If borrowed from an English variety with glottalling of syllable-final voiceless plosives, probably London in view of the work in which they are attested (Kanagaki [1872-76] 1980), it is possible to analyse oorai and ooraioorai as having undergone adaptation of the glottal stop l'?l --7 0 (27).
Chapter 3. Phonology
haNkeCiifu haNkaCiifu
'ibid: 'ibid:'
[0 [0
Eng. yard
yaaru yaado yaarudo
'yard (of cloth)' 'yard (in golf)' 'yard'
[•1884] ,.1878] ,.1869]
Eng. board
booru 65 boodo
'cardboard' 'ibid:
[0 [0
Fr. croquette
koroQke koroQkeQto
'croquette' 'ibid:'
[0 1909] [•1919] 66
Eng. ballast
barasu barasuto
'ballast' 'ibid:
[0 [0
1912] 1899]
Eng. cement Du. cement
semeN semeNto
'cement' 'ibid:'
[0 [0
191<1] 1811]
Eng. waste
uesu uesuto
cotton waste, rag waste' 'waste'
[0 [0
1926] 1930]
Eng. jitterbug
jiruba
'jitterbug'
[•1947]
Eng. astringent
asutoriNzeN asutoriNzeNto
'astringent' 'ibid:
[0
Eng. check it out
(:ekira
'check it out!'
[n/a]
1886] .1887]
1897] 1914]
1953] ,.1955]
The final category is that of donor ltJI and illustrated in (176). As shown in (70) above, English donor lrJ I ~ /Ng/ when not followed by a velar or alveolar plosive. 'Ihe examples in (176) have undergone deletion of the /g/ segment and are most likely auditory loans from speakers of English varieties where lnl is f(mnd in place of IIJ I· In this category full fixms are less evident. SaafiNgu, daar· iNgu and buumeraNgu can be found, while poliNgu [+1832, from Du. pudding], poQdeNgu [+1872], puQciNgu [+1884], puQjiNgu [+1884] and pudiNgu [0 1912] are all attested.
(176)
Eng. pingpong Eng. boomerang Eng. pudding Eng. tongue Eng. leggings Eng. darling Eng. surfing
piNpON buumeraN puriN taN regiNSU daariN saafiN
'table tennis, pingpong' [0 1905] [ 0 1908] 'boomerang' [•1917] 'creme caramel' 'tongue (meat)' [ 0 1930] 'leggings> [ 0 1948) [•1958] 'darling' ,.1964] 'surfing'
65. Now found only in compounds: e.g. the hybrid daNbooru 'cardboard: 66.
Source, but no attested form, is cited.
125
126
Loanwords in Japanese
Many of the examples of deletion illustrated in (173-175) were first attested in the late 19th century, the period offoreign advisors (§2.4.4) and higher than average auditory contact. Several examples exhibit adaptation of donor ltl or ldl ~ /ri, not surprising in auditory loans: Eng. pudding~ puriN; Eng. jitterbug~ jiruba, Eng. check it out~ cekira. Most of the examples in the final category (176), however, were borrowed comparatively late. It is possible saajiN does not have an auditory source, but rather an orthographic one. The Beach Boys' Su1jin' USA was released in Japan in 1963 as saajiN yuuesuee (Nakayama 1997: 32-36). Note the apostrophe in the English song title. In some instances, ascertaining whether deletion has taken place or not can be highly problematic. Example (177) is a case in point. The full forms (a), (b) and (c) are probably dictionary loans from English. Are the forms (d) and (e) auditory loans exhibiting word-final consonant deletion of (b) and (c), respectively? Or are they simply dictionary loans from French? (177)
3·3
restaurant
a. b. c. d. e.
resuturaNto resutooraNto resutoraNto resutooraN resutoraN
'restaurant' 'ibid: 'ibid: 'ibid: 'ibid:
[0
1906] [*1907] [•1908) [0 1908] [•1908]
Suprasegmental issues
Tokyo Japanese, a sociolect of which is the basis for standard Japanese, is a pitch-accent dialect. 67 Each mora in a word has either a high (H) or low (L) pitch. A word's pitch pattern is predictable providing the location of the accent (henceforth indicated in WIM print) is known. A word may or may not possess an accent. In unaccented words, all moras are H, except for the initial mora, which is L (178a). In accented words, only one accent is possible and this accent is carried by the syllable (McCawley 1978: 119; Kubozono 1996: 71--72). Which syllable carries the accent is unpredictable. Accent-bearing light syllables are accented on their only mora; accent-bearing heavy syllables on their
67. Not all Japanese dialects are pitch accent, but most are. There exist a variety of different pitch accent types. Henceforth, any discussion of pitch accent in thL~ volume should be presumed to be that of standard 'Ibkyo Japanese, data on ~~hich is taken from NHK (1998). See Okuda (1975), Haraguchi (1999) or Ma1tin (1987: 138-161) for detail on dialectal variation in accent.
Chapter 3. Phonology
initial mora. 68 This accented mora isH, after which the pitch falls: all succeeding moras are L (178b--d). All moras before the accented I::l mora are also I::l, except for the initial mora which is L (l78bc), unless it is the initial mora itself \·\'hich is accented (178d). If the initial mora belongs to a non-accent-bearing heavy syllable (178e), then it is L only when a word 'is uttered in a slow, careful, unnatural fashion' (Haraguchi 1999: 7). Thus, in 'natural speech' (178e) is HHH; in 'unnatural fashion' it is LHH (see also Hattori 1954: 246; Martin 1970:429). In the examples in (178), and in all subsequent examples in this section, a dash indicates a mora boundary, while a dot indicates a syllable boundary, by definition also a mora boundary. (178)
b. c. d..
a.
u.ryo-o.ke-i yull-o.ryo-o ya.ka.ma.l-i !1-ta.pa.ta
e.
0-N . •
LHHHH LHLLL LHHHL HLLL HHH NLHH
'rain gauge, pluviometer' 'shipping charges' 'noisy 'fluttering, pattering' 'woman' )
There exist n+ 1 possible pitch patterns for a word with n syllables. This is illustrated in (179), where examples are restricted to nouns containing one, two and three light syllables (thus one, two and three moras). In order to highlight the difference between an unaccented noun and one accented on its final syllable, each is given together with the enclitic nominative particle ga. (179)
lf.i: llga
kaga 2p: •.ciga
fuJIIga bu.taga 3f.1: g.mi.ji ga
jiD.so ga ya.su.llga su.mi.re ga
HL LH HLL LHL LHH HLLL LHLL LHHL
Ll-HH·I
'vinegar' 'mosquito' 'speckle' 'winter' 'pig' 'maple' 'office' 'holiday, rest' 'violet'
The above is an extremely brief outline, using non-gairaigo examples, of what is a complex and heavily researched topic. 69 Accent in the gairaigo stratum exhibits interesting peculiarities and it is to these I now turn.
68. 'Ihus the mora consonants /N Q/ cannot be accented. See, however. Vance (1987:81, 2008a: 147). 69. For more detail on the accent of standard Japanese, see the general accounts in McCawley (1977), Haraguchi (1977, 1999),Vance (1987:77-1 07,2008a: 142-198) or'Isujimnra
127
128 Loanwords in Japanese
In a survey restricted to 7,937 trimoraic nouns, Kubozono (2006: 1141) and Kubozono & .Fukui (2006: 41) showed that 71% (n :::: 2220) of native Japanese and 51% (n :::: 4939) of Sino--Japanese trimoraic nouns were unaccented. The same survey; however, showed that only 7% (n = 778) of trimoraic gairaigo were unaccented. The motivation behind this eschewal of unaccentedness in gairaigo is problematic and the data wide-ranging: the interested reader is referred to Kubozono (2001 b: 119···123, 2006: 1156···1160). Here I will touch only on what is perhaps the most notable finding, the fact that a range of studies (Tanaka 1996, cited in Kubozono 2001b; Kubozono 2001b: 120-121, 2002a:93, 2006: 1159; Kubozono & Fukui 2006: 42···44) have shown that a full 80---90% of quadrimo·· raic loanwords buck the gairaigo trend and are unaccented - provided they fulfil two conditions. As illustrated in (180), these are that the quadrimoraic loan end in two light syllables and that its final vowel not be epenthetic /u/. In (180) and in further examples below d· indicates a light syllable and d' a heavy syllable ( 14, p. 75). (180)
Sp. hiniesta Du. alkali Eng. A.rizona
ao-o-a UO'O'c}
ao-o-a
e.ni.si.da a.ru.ka.ri a.ri.zo.na
'broom (plant)' 'alkali' 'Arizona'
Turning now to accented gairaigo, which comprise the bulk of the stratum, we find word length to be a critical factor in accent placement. Bimoraic gairaigo are accented on their initial mora: (181)
Eng. bar Eng. off Du.gas
11-a lfu !5su
'bar'
'off' , gas (
For accented trimoraic or longer gairaigo, the most common location for the accent is the antepenultimate syllable. According to Kubozono (2006: 1147 ..·1150), this is the case for 96% of accented trimoraic gairaigo and 70% of accented quadrimoraic gairaigo. This is illustrated in (182), where examples are drawn from a range of donor languages spanning the various gairaigo historical phases described in Chapter 2. Recall that accent ..bearing heavy syllables are accented on their initial mora. If, therefore, the antepenultimate mora is the second mora of a heavy syllable, then the accent is shifted one mora towards the beginning of the word.
(2007:66-85). For discussion of accent in compounds, see McCawley (1968: 157-172), Kubo7..(mo ( 1995b, 1999b, 200 I b, 2002a, 2006) or Kubozono & Fnjiu.ra (2004) .•
Chapter .3. Phonology
(182)
[+1587] [ 0 1610] [0+1713] [•1802] [•1804] [0 .1807] [+1850] [0 1877] [•1880] [•1909] [•1917] [ 0 1924] [ 0 1939] [•1948] [•1961] [•1975]
Por. cristao Por. mirra
'Catholic, Christian' ki.riJ.ta--N 'myrrh'> 'mummy' ·-i.ra 'camphor' Du. kamfer 11-N.fu.ru 'catheter' Du. katheter ka.l-e.te.ru 'reindeer' Ainu tonakkay to8ka-i 'balalaika' Ru. 6ananai1:Ka lbalalajkal ba.rall-i.ka 'hotel' Eng. hotel ll.te.ru 'gramme' Fr.gmmme l!lra.mu 'dressing, gauze' Ger. Gaze 11-a.ze sa.moaa.ru 'samovar' Ru. caMoBap Isamovar! 'creme caramel' Eng. pudding ll.ri-N Ger. Rucksack ryu-Q.ku··-Q.ku 'rucksack' 'fried rice' Bei. chaofan jz.I>fPt 1--a.ha--N 'pickled cabbage' Kor. kimch'i ~ ~1 l.mu.ci 'green pepper' Fr. piment 1-Lma-N 'pasta' It. pasta 11-su.ta
With accented trimoraic loanwords, one major instance where the accent typically does not fall on the syllable containing the antepenultimate mora is with words whose syllable structure is light--heavy (ad') and where the light syllable coda is an epenthetic vowel (Tanaka 1996, cited in Kubozono 200lb; Kubozono 1996:75-76, 2001b: 117-119. 2006: 1147; Tanomura 1999:71-72). Here, as illustrated in (183), the accent frequently falls on the final syllable, i.e. the penultimate mora. Compare the examples in (183) with the trimoraic gairaigo in (184). These have the same 6'0' syllabic structure but their light syllable coda is an original, not epenthetic, vowel. 1heir accent location is typically. as expected, in the syllable containing the antepenultimate mora. (183)
(184)
Eng. ski Eng. crew Eng. spin Eng. grey Eng. rally Eng. dummy Eng. ribbon Eng. buzzer
aa aa iJ<j
su-1-i
kuJI--u su-JI-N gu.e
M M M
l.ri-i
aa
bo-N • .za-a
M
~mi-i
'ski(ing)' 'crew' 'spin' 'grey' 'rally' 'dummy' 'ribbon' 'buzzer'
At first glance, it may appear that the motivation behind the examples in (183-184) is that they keep the original accent/stress of the donor word. As Kubozono (2006: 1147) and others have pointed out, though, donor accenti stress is not necessarily preserved elsewhere: e.g . • nu-u ~Eng. ca.n6e, Jra-N ~ Eng. I1·&n. Sibata (1994:408) and Mutsukawa (2009:43---46) have noted that 35%
129
130
Loanwords in Japanese
(297/858) and 31% (3·11/1090), respectively, of gairaigo borrowed from English do not have the same accent/stress location as their donor word. Nevertheless, others, such as Giriko (2006), claim original accent/stress plays a larger role. With quadrimoraic or longer accented loanwords, a major exception to the antepenultimate rule is found, once again, among words whose final two syllables are light-heavy ((r(j) (Katayama 1995, 1998: 179---185; Suzuki 1995; Shinohara 2000:81--83; Kubozono 2002a:90, 2006: 1153---1156). Here, the accent frequently falls (and with new borrowings and innovative pronunciations is exhibiting an ever growing tendency to fall) on the syllable containing the pre-antepenultimate mora (fourth from the end of the word): (185)
3·4
aaa Eng. policy 11.ri.Si-i Eng. cosmopolitan MMJO' ko.su.mo. .ri.ta-N aaa Eng. junction 11-N.ku.so-N Eng. Westminster MMM we.su.to. • N.su.ta-a
'policy' 'cosmopolitan' '(motorway) junction' 'Westminster'
Mora-dipping
Truncation or compression is common cross-linguistically. In Japanese it is not confined solely to the gairaigo stratum and takes many forms. These can be broadly divided into three types: mora-clipping, compound reduction and Roman alphabet acronyms. While mora--clipping may be f(mnd with any word f(1rm, compound reduction and Roman alphabet acronyms are found only with compounds. Since compound reduction is a morphophonological phenomenon I will examine it in the following chapter (§4.3.1). Roman alphabet acronyms are an orthographic phenomenon and will be considered in §5.4.2. Mora-clipping will be treated in this section. All three types of truncation are summarized in Table 3.9 with the gairaigo compound koNbinieNsusutoa 'convenience store' as input. In practice, if a word undergoes truncation at all, it is typically frmnd in one truncated form only or, if multiple truncated forms do exist, then usage is heavily skewed towards only one of them. In the case of koNbinieNsusutoa, the back-clipped form koNbini is found with overwhelming frequency (.r), although truncated forms may also occasionally (I:EI) be found which employ fore--clipping, compound clipping, ellipsis and even other forms not noted here. A truncated form (e.g. the back-clipped koNbini) can undergo further truncation (e.g. to fore-clipped bini) and even subsequent verbalization (e.g. biniru 'go to a convenience store', cf. (201) on p. 139). A cross (;c) in the rightmost column of Table 3.9 does not necessarily mean a truncated form does not exist, rather that the author is unaware of one. Truncation processes are
Chapter .3. Phonology
highly fluid and often confined to jargon, youth speech, slang and dialect. None of the truncation processes show-n in Table 3.9 are restricted to gairaigo, although they occur a great deal more commonly in this stratum. Table 3.9. Gairaigo truncation processes Truncati.on moraclipping
Input Section (examples)
back-dipping
fore-clipping
§3.4 (186-192, 195-196)
koNbinieNsusutoa koNbinieNsusutoa
§3.4 (193)
koNbinieNsttsutoa koNbinieNsttsutoa
§3.4 (194)
koNbinieNsusutoa koNbinieNSUSUfoa
Output
Found
koNbini koNbi
./ )l
etc ... toa sutoa
)l
liD
etc ... mid-clipping
koNn!su koNeNtoa
)l )l
etc ... compound compound cHpp'lng redtKiion ellipsis portmanteau formation Roman alphabet acronyms
§4.3.1.1 (219-229) §4.3.1.2 (230---233) §'1.3.1.3 (234) §5.4.2
koNbinieNSU ·I·SUtoa koNsuto lmNbinieNsu+sutoa koNsu
etc ... hmbinieNSU+SUtoa sutoa
liD )l
liD
koNbinieNSU+ sutoa koNbinieNSU
)l
koNbinieNsu +Sutoa koNa
)l
convenience store
)l
silesu ((CS))
(Table 5.5)
Mora-clipping differs from deletion (§3.2.3) in that it occurs after all the loanword adaptation processes detailed in §3.2 have applied. A donor word may thus, in theory, undergo deletion during adaptation and then suffer further mora-dipping thereafter. Within deletion itseli: the subcategory of donor word-final consonant deletion (175) can be especially difficult to differentiate from back-dipping. Deletion is typically the product of an auditory source (Figure 3.3), with word-final consonant deletion mirroring colloquial donor pronunciation. On the other hand, mora-clipping occurs post-adaptation and operates independently of any auditory loan process. Moreover, back-clippings where only a single mora is lost are rare. The primary motivation behind mora--clipping (and truncation processes in general) is the above-average word length (compared to the Japanese lexicon as a whole) of gairaigo, a circumstance brought about in particular by epenthesis (§3.2.2). The more moras a gairaigo contains, the more likely it is to undergo mora-dipping. In a survey of 6,249loanwords Kanno (1985: 56-57) f(mnd that, while approximately
131
132
Loanwords in Japanese
35% and 25% of 11 and 10 mora gairaigo have clipped forms,7° the proportions fall to 10%,7% and 3% for 5, 4 and 3 moragairaigo, respectively. In the discussion which follows, I ignore gairaigo such as iN/oro 'intro: basu 'bus; neQto '(inter)net', demo 'demO, aNpu 'amP, funmku 'frank(furter)', maiku 'mic/mike; etc. which are, in all likelihood, adapted from forms already clipped in the donor language.
Mora-clipping in loanwords can be divided into three categories: back-clipping (apocope), fore-clipping (aphaeresis) and mid --clipping (syncope). The first of these clipping processes, back-clipping, is, by a considerable margin, the dominant one. In mydatabase 71 of34J mora-clippedgairaigo the back-fore-mid ratio is 85:8:6. 'Ihe motivation f()r a given gairaigo to either back-, f<)re- or mid-clip remains obscure, however, and further research is required. In back--clipping, the latter portion of a gairaigo is clipped and only the first two (186), three (187), four (188) or, occasionally, five moras retained. Retention of the first two moras - bimoraic back-clipping - is slightly preferred, making up 41% of my database, while trimoraic and quadrimoraic back--clipping account for 33% and 25%, respectively. It6 (1990: 217) claims that, except f()r a very few written abbreviations whose status is questionable, monomoraic back-clippings are impermissible, as are those composed of five moras or more. While there are no monomoraic back-clippings in my <.latabase, there are four examples with five moras: apoiNto 'appointment', koNpaato 'compartment; (189) koNseNto and (195) erekiteru. The first two of these have English sources ending in the morpheme ment, on whose behaviour see Pootnote 73. (186)
Eng. chocolate Eng. ecology Eng. building Ger. Seminar Eng. (cash) register Fr. millimetre Eng. connections
cokoreeto ekorojii birudiNgtt zeminaaru rejisutaa mirimeetoru konekusoN
> > > > > > >
coko eko biru zemi reji miri kone
'chocolate' 'green, eco-friendly' 'building' , 'seminar 'till, cash register' 'millimetre' 'personal contacts'
While Kanno makes no distinction between mora-dipping and compound dipping (§4.3.1.1) in his statistics, they still serve as a useful indicator of the relationship between length and dipping in general. 70.
All database tokens appear as either an entry (or as a redirection to an entry) in the Wikipedia Japan pages during 2009 or 20 l 0, or in at least one of the following ll dictionaries: Umegaki (1956), Arakawa (1977), JKS (1977, 2003,201 0), Maruyama et al. (1992), Yonekawa (1997), GJH (2001), Watanabe et al. (2003), Shinmura (2008), SSDH (2010).
71.
Chapter 3. Phonology 133
Eng. mistake
Fr. mayonnaise Eng. strike (187)
misuteeku mayoneezu sutoraiki
Eng. permanent (wave) Fr. centimetre Ru. KaMTiaHHJ'I lkarnpaniijal Eng. •:mimation Eng. supplement Ger. Impotenz Eng. sartdv.-ich Fr. esthetique Eng. television
Eng.l>.-facDonaid's (188)
Eng. inflation Ru.lintielliig)encijal Eng. eucalyptus Eng. accelerator Eng. rehabilitation Eng. infrastructure Eng. asparagus Eng. extension Eng. restructuring Eng. automatic
> misu 'error, slip, mistake'
> mayo 'mayonnaise' > suto '(labour) strike'
paamaneNto paarna seNCirneetoru > seNci kaNpania kaNpa anirnedoN anime sapurirneNto sapuri iNpoteNCtl > iNpo saNdoiQci saN do esutetiQku esute terebijoN terebi makudonarudo > makudo 72
iNfuredoN iNterigeNca yuukariputasu akuserureetaa rihabirlteesoN iNfurasutorakucaa asuparagasu ekusuteNsoN risutorakui::uariNgu ootomaciQku
> > > > > > > > > >
iNfure iNteri yuukari akuseru rihabiri iNfura asupara ekusute risutora ootoma
'perm' 'centimetre' 'fund -raising campaign' 'cartoon' 'vitamin supplement' 'impotence' 'sandwich' 'beauty care, beauty salon' 'TV' 'MacDonald's' 'inflation' 'intellectual' 'eucalyptus> 'throttle, gas' '(medical) rehabilitation' 'infrastructure' 'asparagus' '(hair) extension' , 'downsizing 'automatic (transmission)'
Although, as we shall see in the following chapter, the form of reduction usually favoured in compounds is t}'11ically compound clipping (~.3.1.1) or ellipsis (§4.3.1.2), back--clipping may occur here also. Some examples are illustrated in (189), where'+' indicates a word or element break. (189)
Eng. convenience store Eng. department store Eng. metabolic syndrome Eng. concentric plug
koNbinieNsu+sutoa > koNbini > depaato depaatomeNto+sutoa metabo metaboriQku+siNdoroomu > koNseNto koNseNtoriQku+puragu
'convenience store' 'department store' 'being overweight' 'plug, socket, outlet'
Analyses by lt6 (1990), It6 & Mester (1992) and Labrune (2002) have demonstrated that the patterns of back-clipping which arise in f()rms such as those in (186---188) are governed by patterns of light (cr) and heavy (a) syllables (14, p. 75). To an overwhelming degree, the output of the back-clipping process is restricted to five forms: era, crcrcr, 6aM, ocr and Mer. Thus, heavy syllables are typically confined to
72. Used mainly in the Kansai region. In Kanto and flllther north, the typical tmncation is maQku (Nagase 1999: 19-21). Here, a compound- biQgu+maqku 'Big Mac: maQku+seeku 'Me Shake' or the like- has undergone ellipsL~ (§4.3.1.2).
134 Loanwords in Japanese
initial position in trimoraic (aa) and quadrimoraic (aaa) forms: e.g. (187) paama, seNci, kaNpa, etc.; (188) iNjure, iNteri, yuukari, etc. Back--clippings consisting of a single heavy syllable (e.g. *paa, *seN, *kaN, *iN, >~yuu, etc.) or ending in a heavy syllable (e.g. *t'okot-ee, *animee, *koNseN, etc.) are generally not tolerated.'73 Given that aa, Ma and aaaa are all acceptable outputs of the back-clipping process, what then motivates a loanword such as (188) tJ('}-()'t'J(t(j akuserureetaa to undergo back-clipping to aaaa akuseru, rather than to aaa "akuse or aa >l-aku? Given both d'cr and d'crcr are acceptable outputs, what motivates a loanword such as (188) d'acrd'cr iNterigeNca to undergo back-clipping to d'acr iNteri, rather than to ocr *iNte? Motivation appears to lie in Labrune's (2002: 106) claim that, in general, loanwords are back-clipped immediately before the This is illustrated in (190). After back-clipping, the location of the accent may move or the clipped form may become unaccented. (190)
Eng. accelerator Eng. terrorism Ru. lintie!Vigiencijal Eng. symposium
60'66 ... akusentttaa > 6M
M Ito 0'60'... iNteri[INca > aM iNteri 'intellectual'
aM...
SiNpolllumtt
>
aa
INro
'symposium'
Of course, Labrune's claim does not apply in cases where a gairaigo is unaccented. Nor, since back-clippings composed of more than five moras are illicit, does it apply in cases where a gairaigo is accented seven moras or further from the beginning of a word. The output of back-clipped gairaigo accented on the first or second mora is governed by rules different from those in (190). Since monomoraic back-clippings are illicit, those accented on the first or second mora will have a bimoraic output. This is illustrated in (l9lab ). If, however, such a gairaigo begins in a heavy syllable, then a bimoriac output is blocked by the constraint against back-clippings consisting of a single heavy syllable. 1hus, Lab rune claims, a gairaigo accented on its initial mora and beginning in a heavy syllable74 will be clipped to three moras. This is illustrated in (191cd).
73-
In my database, 96% of back-clippings having four moras or less are of the form O'Ct,
aaa, ctetetct, aCt or acta, thus corroborating previous research. Common exceptions include purezeNteesoN > purezeN 'oral presentation' and paareNteeze > paareN 'parenthesis' (Ger. Parenthese). Other exceptions, from loans ending in -meNto ~----Eng. -ment, cluster around a Met output. These include apaato < apaatomeNto ~--- Eng. apartment and (189) depaato (see also Tanomura 1999: 82). If, rather than back-clipping, the patterns found in these loans are viewed as morphological reduction (§4.2), i.e. deletion of the English morpheme rnent, then these exceptions are eradicated. 74· §3.3.
Words accented on the second mora and beginning in a heavy syllable do not exist. See
Chapter .3. Phonology 1.35
Here, the constraint against monomoraic back-clippings disallows *ka and *me, while the constraint against single heavy syllable back-clippings rules against *kaa and *meN. (191)
a. b. c. d.
u"' ••• Eng. festival Stti~-~ru <1' ... Eng. ecology 0)11 a ... !ladegaN Eng. cardigan Eng. maintenance cr ... llltaenaNsu
> lsu > lko > !atde > .Nte
'festivar 'green, ec.o-friendly' 'cardigan 'maintenance'
There are, nevertheless, a significant number of exceptions to Labrune's claims in (190--191). She states these are overwhelmingly confined to loans accented on the third mora or beyond and contends (opus cit.: 111-114) they can be largely explained by constraints against the back-clipped output ending in an epenthetic vowel (192a) and, to a lesser extent, an /rV/ mora (192b). In (192a), the expected back-clipping *iNfu. is not found, Labrune claims, because the /u/ is epenthetic (§3.2.2.1). Note, however, that the attested form iNjure ends in an /rV/ mora which she also claims is avoided. With (192b), the accent on the second mora, together with the fact that the initial syllable is light, mean that we expect the bimoraic back--clipping *Jura (ct~ 19lb). This is not found, Labrune maintains, because of the avoidance of a final /rVI mora. (192)
a. b.
Eng. inflation iNf.soN > iNfure (*iNfu) 'inflation' Eng. fraction iliJ.usoN > furaku (*fura) 'fraction'
Fore-clipping is the opposite of back-clipping and is considerably less common. The final two (8/28, 29% in my database), three (12/28, 43%) or four (8/28, 28%) moras are retained, never the final one and never the final five or more. Some examples are shown in ( 193). Fore-clippings are too few to allow a judgement to be made on whether the location of the accented mora plays a role in the fore-clipped output. (193)
wanisu Du. vernis Eng. propeller puropera Eng. platform puraQtohoomu arubaito Ger.Arbeit Eng. triangle toraia.lll guru Eng. velveteen berubeQciN
> > > > > >
nisu pera hoomu baito aN guru beQCiN
'varnish' 'propeller' 'railway platform' 'part-time job' 'menage atrois' 'velveteen'
Mid-clipping is the rarest of the three types of mora-clipping. Here, moras from anywhere in the full form are clipped, typically including one or more word-final moras. Some examples are shown in (194).
136
Loanwords in Japanese (194)
Du. morfine Eng. entertainment Eng. instructor Eng. transparency Eng. correspondence
moruhine eNtaateeNmeNto iNsutorakutaa toraNsupeareNSii koresupoNdeNsu
> mohi > eNtame > iNtora > torapeN > korepoN
'morphine' 'entertainment' 'instructor' '(OHP) transparency' 'correspondence'
TI1ere are no attested examples of mora-dipping, of any t}'lJe, from the Iberian borrowing period. The earliest instances date back to the period of Dutch borrow·· ing (§2.3), the oldest mora-clipped forms of which I am aware being: (195)
[•1779] Du. elektriciteit erekiteruseeriteeto > erekiteru 'electricity' [•1798] > ereki 'electricity'
Finally, just as with other adaptation processes outlined in this chapter, moraclipping can lead to homonymy. A few examples are shown in (196). See also puro in Table 2.11 on p. 60. (196)
Eng. diamond Eng. diagram Eng. install Eng. instructions Eng. instrumental
daiyamoNdo daiyaguramu iN sutooru iNsutorakusoN iNsutorumeNtanl
> daiya > daiya > iNsuto > iNsuto > iNsuto
'diamond' '(train) schedule' 'installing (software)' 'instructions' 'instrumental'
CHAPTER
4
Morphology, tnorphophonology and semantics
Loanword morphology and morphophonology encompass three main areas. 'Ihe most fundamental of these, examined in §4.1, is the question of how loanwords inflect once borrowed into Japanese. In some cases, a gairaigo undergoes morpho-logical reduction during the borrowing process: this I consider in §4.2. The most heavily researched area of loanword morphology is the range of phenomena associated with compounds: these will be dealt with in detail in §4.3. Finally, the semantics of loanwords will be discussed in §4.4. Throughout this chapter, a '+' symbol indicates an element or word break.
4.1
Loanwords and morphemes, loanwords as morphemes
Once borrowed, most gairaigo function as nouns. In the NINJAL (1964: 57) magazine survey (§1.3), 95% (2820/29M) of gairaigo types were nouns, whereas only 78% (23783/30331) of types were nouns across all vocabulary strata. As nouns are not declined in Japanese, most gairaigo therefore exhibit no mor-phology. 'Ihis is not the case, however, for those gairaigo which function as verbs, adjectives and adverbs. Loanwords functioning as verbs also function as nouns and these nouns exhibit two radically different verbalization patterns. The overwhelmingly dominant pattern is the use of the verbalizer sunt 'to do': (197)
Eng. up Eng. propose Eng. sign Eng. announce Eng. counselling
aQpu puropoozu saiN anauNsu kauNseriNgu
(rise:!' increase~ 'proposal' 'signature) 'announceme11t' 'counselling'
aqpusuru puropoozu suru saiNsuru anaur,rsu suru kauNseriNgu suru
.:increase' 'propose marriage' 'sign' '"annowice'
'counsel'
\\'hether a loan noun such as anau.Nsu is from the English verb announce or a morphologically reduced (§4.2), or perhaps back-clipped (§3.4), form of the
138 Loanwords in Japanese
English noun announcement is moot. 75 \A/hichever was the case, anauNsu was then verbalized, yielding anauNsu suru 'to announce'. This pattern of verbalization is identical to the dominant pattern found in the Sino·· Japanese stratum: e.g. koo.~oo 'negotiation' and koosoo suru. 'negotiate'. The second, though highly marked, verbalization pattern exhibited by a small number of gairaigo nouns is the suffix ·-ru. This suffix may be attached directly to the noun itself (198) or, since the favoured length is 3 or 4 moms, to a mora·· clipped (§3.4) form thereof (199). (198)
(199)
Eng. memo memo Eng. panic paniQku Eng. demo demo Du.gas gasu dafu Eng. duff Fr. sabotage Eng. Starbucks Eng. Febreze Eng. Wikipedia Eng. harmony
'memo, note' 'panic' 'demo' 'gas' 'duff (in golf)'
sabotaaju sut.aabaQkusu faburiizu wikipedia ha.amonii
> > > > >
memoru pani (Q)kuru demoru gasuru dafuru
'work-to-rule, idleness' 'Starbuck.~'
'Febreze' 'Wikipedia' 'harmony'
saboru sutabaru faburu > wikiru > hamoru
'jot down 'panic' 'go on a demo' 'mist up' 'duff' 'skip classes' 'go to Star bucks' 'deodorize' 'search for on \'llikipedia' 'harmonize'
In most cases, verbs such as those in (198-199) have alternative forms with the verbalizer suru (e.g. paniQku suru). Although native Japanese verbs ending in -ru may be either consonant or vowel verbs, the two conjugations to ""·hich all modern Japanese verbs belong, gaimigo verbalized in the manner of (198-199) conjugate without exception as the former. In other word.s, they are treated as having a stem ending in r.. to which verbal suffixes are attached: e.g. wikiranai 'not search for on \Vikipedia: not "wikinai. If a gairaigo noun itself end.s in -ru then it may function as a verb without a -1·u suffix. It too will conjugate as a consonant verb: (200)
Eng. Eng. Eng. Eng.
Google double trouble cycle
guuguru daburu toraburu saikuru
'Go ogle' 'double' 'trouble' '(bi)cycle'
> guuguru 'Go ogle' > daburu 'overlap, be duplicated' > toraburu 'cause trouble' > saikuru '(bi)cycle'
Only a very small proportion of gairaigo nouns ending in -1·u are verbalized. as in (200). Most, for example risaikuru 'recycling' or koNtorooru 'control: are verbalized as in (197): risaikuru suru and koNtorooru suru. Furthermore, even those few gairaigo which do verbalize as in (200) typically possess competing forms with suru: e.g. guu.gu.ru suru..
75· As, cross-linguistically, nouns appear to be borrowed more easily than other pa1ts of speech (Moravcsik 1978; van Hont & Muysken 1994), the latter is more likely.
Chapter 4. Mmphology, morphophonology and semantics 139
The earliest written attestation for the -ru verbalization pattern illustrated in (198···200) is 0 1925 for saboru, but most are considerably more recent, typically found in slang and youth speech, and are highly transient. Listed in ( 201) are (were) examples of the more ephemeral. See Yonekawa (1997) for further examples. (201)
Eng. neglect Eng. convenience st
negurekuto koNbini 7.eroQkusu parodii
'neglect' > neguru 'convenience store' > biniru zeroru 'photocopy' 'parody' > paroru
'neglect' 'go to a convenience store' 'photocopy' 'parody'
Many loanwords also function as adjectival nouns (nominal adjectives). When used attributively, these modify nouns by means of na, a form of the copula (202ac). When used predicatively, they use the copula (202bd). (202)
a. b. c. d.
Eng. severe sibia
'harsh reality' 'reality was harsh' 'the best way' besuto besuto na hoohoo > kono hoohoo ga besuto ja nai 'that way isnt best' sibia na geNjiCU
> geNjicll ga sibia
Eng. best
Only a very small number ofloanwords function as true adjectives. Extremely rare and highly marked, they employ the morpheme-ito indicate the non-past, as illustrated in (203), as well as other conjugational endings: e.g. non-past negative piNkukunai 'is not pink past piNkukaQta 'was pink etc. All the examples in (203) also exist as adjectival nouns: e.g. piNku na koneko 'pink kitten'. (203)
Eng. grotesque Eng. pink Eng. erotic Eng. now
gurotesuku piNkll eroCiQku nau
> > > >
guroi piNkui eroi naui
'is grotesque, bizarre' 'is pink, erotic' 'is erotic' 'is trendy, in' 76
Adjectival nouns like those in (202) may also function as adverbs by means of ni, another form of the copula: e.g. sibia ni 'severelY: The true adjectives in (203) use the morpheme -ku: e.g. guroku 'grotesquely'. ,-.._,
Other morphological affi.xes, both prefixes (204) and suffixes (205), may be attached to gairaigo. These include, but are by no means restricted to, the prefixes
76. Although uaui as an adjective has long pas:;;ed its sell-by date, nau has made a comeback in severely character-restricted "J\vitter tweets, with the meaning'to be ... now: 'Ihetwo characters required to write nau (usually tr::? ) make it preferable to the three needed for ui iru ( (.c:_ \r' .'if,), orfourfor ui imasu UC:vYj:·<J): e.g. eki 1wu <"-:ld:?) for eki ni iru (f~t·C:v' 0) 'I'm atthe station:
140
Loanwords in Japanese
gyaku- 'reverse, retro-: haN- 'anti-, counter-: haN- 'semi-: koo- 'high-: the honorific O·, sai·· 're--'; and the suffixes -jiN 'person: -ka '--izatiml, the plural ··ra, -saN 'pro·· duced, grown; ··SCi '-ness, ··ity; ·Sei 'manufactured: -.Siki '··Style, lcl, -Jugi '··ism.' and -yoo 'for the use of'.
a
(204) Eng. course Eng. neutrino Fr.jupon Ger. Energie Por. tabaco Eng. open
koosu nyuutorino zuboN enerugii tabako oopuN
> > > > > >
gyakukoosu haNnyuutorino haNzuboN kooenerugii otabako saioopuN
'reverse course, reactionary' 'antineutrino' 'shorts' 'high -energy' 'cigarette' 'reopening (of store, etc.)'
(205) Fr. Suisse
suisu surimu oobii oosutoraria a.rukari amerika furaNSU mamkusu maQku
> > > > > > > > >
suisujiN surimuka oohiira oosutorariasaN arukarisei amerikasei .furaNsusiki. marukususugi maQkuyoo
'Swiss (person)' 'downsizing' 'alumni, ex-employees, etc: 'gmwn in Au'>tralia' 'alkalinity' 'made in the us~ 'French-style' 'Marxism' 'for use with M.acs'
Eng. slirn Eng.OB Eng. Australia Du. alkali Eng. America Fr. France Ger. (Karl) Marx
Eng ..Mac
(........,
Gairaigo may themselves function as affixes. Th.ese include, and once again are by no means restricted to, the prefixes in (206) and the suffJXes in (207). Where an affix can be attached to another gairaigo, examples have been restricted to those where the resulting word does not exist in the donor language \·\'ith the same or similar meaning. All affixes illustrated below are borrowed hom English and a few of them have undergone semantic change: mai means 'one's own' and can be used of a second or third person; the currently very much in vogue sirubaa means 'old age, OAP, senior citizen'; while -aa has undergone semantic broadening to 'someone who is into'. (206) hai maikuro mai maruci mini noN noo oobaa semi sirubaa suupaa
za
high + Eng. teen micro-+ native my + Eng. boom multi- + Sino mini-+ Sino non- + Eng. section rro + Eng. cut over- + Eng. drug semi- + Eng. long silver + Sino super + Sino the+ Du. bier
> > > > > > > > > > > >
haitiiN ma il:uro ha maibuumu maruciniN geN miniyoNku noNsekusoN nookaQto oobaadoraQgu semiroNgu sirubaasaNgyoo suupaadaikaiteN zabiiru
'late teens> 'microwave-' 'personal fad' 'person of many talents' 'small 4WD car' 'non·· specialist' 'uncensored' 'overdose' 'medium length' 'elderly care industry' 'super giant slalom' 'the (only/best) beer'
Chapter 4. Mo.tphology, morphophonology and semantics
(207)
aa aQpu dauN furu iN izumu maN ofu siQpu
Fr. Chane! + er Eng. level + up Eng. cost + down Eng. heart + -ful Eng. season + in native + -ism Eng. gag + man Eng. price + off Eng. skin + -ship
> saneraa > reberuaQpu > kosutodauN > haatofum > SiizuNiN > osareizumu > gyagumaN > puraisuofu > sukiNsiQpu
'someone into Chane! goods' 'improvement' 'reduction in price' 'heart--warming' 'start of the (sports) season' 'chic-ism' 'comedian, comedy writer' 'discount' 'physical contact'
The suffix -aa exhibits unusual behaviour in that it causes elision of the precedingvowel (e.g. saneru + aa > saneraa; also saakuru 'university dub' (Eng. circle) + aa > saakuraa 'someone who is always at their university club: and many others). It may also appear as -raa: e.g. mayoma 'someone who eats mayonnaise with everything'. Reanalysis of English affixes also occurs. Eng. -tic (208) is a reanalysis of -ic motivated by the fact that many donor -ic words end in -tic (e.g. maguneciQku 'magnetic', pedaNciQku 'pedantic: erociQku 'erotic: asurdiQku 'athleticS, etc.). (208)
4.2
CiQku Sino+ --tic > maNgaciQk'U 'manga-style, mangaesque' Ger. Miirchen +-tic > meruheNciQku 'fairytale-like'
Morphological reduction
During adaptation donor words are typically shorn of native morphology. Listed below are selected examples of this process at work with plural (209), past participle (210), gerund (211), possessive (212) and other miscellaneous morphemes (213). Several ditlerent morphological markers may be deleted from the same donor word or compound (214). All further examples in this section are from English unless otherwise stated. (209)
on the rocks ladie.s first three strikes pyjamas sunglasse.s
(210)
smoked salmon salaried man scrambled egg condensed milk mixed grill
(211)
skiing measuring spoon spelling happy endirtg
ONZ<'1roQku rediifaasuto suriisutoraiki pajama saNgurasu
'on the rocks' 'ladies first' 'three strikes' 'pyjamas' 'sunglasses'
sumookusaamoN sarariimaN sukuraNburueQgu koNdeNsumiruku miQkusuguriru sukii mejaasupuuN superu haQpiieNdo
'smoked salmon' 'white collar worker' 'scrambled egg' 'condensed milk' 'mixed grill'
'skiing 'measuring spoon' 'spelling' 'happy ending'
141
142
Loanwords in Japanese
(212)
Valentine's Day bandsman lambswool l'vfadame 'llmaud's
(213)
engagement ring Du. reumatisch 'rheumatic' lucky seventh
(214)
mashed potatoes mixed vegetables women's lib
bareNtaiNdee baNdomaN ramuuuru madamutaQsoo
'(St.) Valentine's Day' 'bandsman, musician' 'lamhswool' 'Madame Tussaud's'
eNgeejiriNgu ryuumaci raQkiisebuN
maqsupoteto miQkusubejitaburu uumaNribu
'engagement ring' 'rheumatism' 'lucky sevent'
'mashed potatoes' 'mixed vegetables' 'women:'s lib'
Nevertheless, there are a significant number of gairaigo where donor morphology remains intact, as illustrated in (215).1here appears to be a strong tendency for more recent borrowings to retain donor morphology, perhaps connected with an increased general awareness of English grammar among Japanese speakers. That said, there are a number of older borrowings (e.g. arupusu ~<\Ips' 0 1893) which have also retained donor morphology. Further research is required in this area. (215)
singles A.lps idling outsourcing informed consent poached egg synchronized swimming
siNgurusu arupusu aidoriNgu autosoosiNgu iNfoomudokoNseNto pooCidoeqgu siNkuronaizudosuimiNgu
'(tennis) singles'
'Alps' '(engine) idling' 'outsourcing' 'informed consent' 'poached egg' 'synchronized swimming'
Wl1en a phrase is borrowed, conjunctions and articles may be deleted. This is shown in (216). In examples (216cd), truncation also occurs subsequent to article deletion. Morphological reduction of conjunctions and articles is not \·\'ithout its analytical problems. For example, the presence and absence of articles in English can differ markedly from variety to variety. The loan oNea 'on air: from Eng. on the air, is sometimes cited as an example of article deletion, but in many English varieties, including the author's, on air without the definite article is perfectly acceptable. Example (216c), too, can occur without an article. Note, however, the interesting eNdoraN 'hit·and .. run (in baseball); where we find deletion of hit instead of deletion of and. (216)
a.
b. c. d.
game and set geemuseQto shampoo and set saNpuu~eQto off the record ofurekoodo > ofureko copy and paste kopilpeesuto > kopi(Q)pe
'game and set (in tennis)' 'shampoo and see 'off the record, between you and me' 'copy and paste (in word processing)'
Finally, in a few cases morphological accretion may occur. Typically restricted to plural morphemes, donor words are assigned a donor morphological marker
Chapter 4. Mo.tphology, morphophonology and semantics 143
during the adaptation process. Although in some cases one can argue it is the plural rather than the singular which has been borrowed, the examples in (217) exclude donor nouns which typically occur in pairs (e.g . .~uuzu 'shoe') or only rarely in the singular (e.g. takosu 'taco'). (217)
4·3
top pop wafer
toqpusu poQpusu uefaasu
'top (garment)' 'pop (music)' 'wafer (biscuit)'
Compound phenomena
In terms of their etymological composition, two different types of gairaigo compound must be distinguished. The first ty11e is the compound borrowed directly from a donor language, illustrated by (218a); the second the compound created in Japan from two independent gairaigo and illustrated in (218b ). (218)
a.
Eng. manpower
maNpawaa
'manpower'
b.
Eng. baby+ Eng. circle
bebiisaakuru
'playpen~
The latter are sometimes called 'made-in-Japan' compounds since, in the Japanese literature, they are typically held to be a subgroup of waseieigo :fP~'L~:-~ 'made-·in ·.Japan English'. This designation is not only unwieldy, itis also vague. In the Japanese literature the term waseieigo may also refer to morphologically reduced English gairaigo words and compounds (§4.2), to English compounds borrm·\'ed directly but \·vhich have been reduced (§4.3.1), and even to English gairaigo words and compounds which have undergone semantic change (§4.4). Throughout this section, therefore, I will refer to compounds such as those in (218b) as 'assembled' compounds, \·vhile employing the term 'imported' compound for those borrowed directly (218a). Since a similar phenomenon is found in other vocabulary strata, 'imported compound' and 'assembled compound' should hereafter be understood to refer to compounds made up entirely ofgairaigo elements. Two major gairaigo compound phenomena are apparent in Japanese: compound reduction (§4.3.1) and sequential voicing (§4.3.2). 4.3.1
Compound reduction
Three strategies ofgairaigo compound reduction are found. The most common of the three is compound dipping (§4.3.1.1) and care should be taken to distinguish this phenomenon from mora-clipping (§3.4). Less common is ellipsis (§4.3.1.2), while a third strategy, portmanteau formation, is rare (§4.3.1.3). See Table 3.9 on p. 131 f(H a summary ofgairaigo truncation processes.
144 Loanwords in Japanese
Compound clipping The vast majority of gairaigo compounds contain tw'O elements. Compound clipping involves reduction of such compounds to anything from lwo to six moras. This is achieved through combining the initiaF 7 moras of each of the two elements of which the compound is composed (It6 1990; Labrune 2002: 102-104). In some cases only one element is clipped: e.g. haioku 'high octane petrof from hai + okutaN, where hai is only two moras long and clipped no further. In the discussion which f()llows, I refer to these as 'half clippings', to those where both elements are clipped as 'full clippings'. Compound clipping is commonly employed in slang, jargon and youth speech and is thus highly ephemeral in nature. Below, I have endeavoured where possible to cite examples in reasonably ·widespread use. Combining the first two moras (fl) of eachofthe two elements to form a 2fl+·2fl clipped compound is the pattern \·vhich is overwhelmingly favoured. This is illustrated for both imported (2l9) and assembled compounds (220). In my database 78 of 455 clipped gaimigo compounds, 76% (344/455) are 2~L+2fl. If half clippings are ignored, this proportion increases to 80% (292/366). 4·3.1.1
(219)
Eng. sexual harassment sekusuaru+harasumeNto waado+puroseqsaa Eng. word processor dejitaru+kamera Eng. digital carr.era rimooto+ kol'tiorooru Eng. remote control Eng. second hand sekoNdo+haNdo kurostt+kal'tiorii Eng. cross cotmtry Eng. mother complex mazaa +ko~pureqkusu Eng. patrol car patorooru +kaa sukeeto+ boodo Eng. skateboard haNgaa+sutoraiki Eng. hunger strike
(220)
Eng. }emls +pants Eng. potato + fty Eng. pocket + monster Eng. resort + lover Eng. engine + stall
> sektib.ara waapuro dejikame > rimokoN > sekohaN kurokaN mazakoN > patokaa sttkeboo haNsuto
jiim,u +paNCU > _iiipaN poteto·t-furai potefura pokeQto+moNsutaa > pokemoN rizooto+rabaa > rizoraba eNjiN+sutooru eNsuto
'sexual harassment' 'word processor' 'digital camera' 'remote control' 'second hand, used' 'cross country' 'mother complex' 'police car' 'skateboard' 'hunger strike'
'jeans' 'chips, French fries' 'Pokemon' 'holiday romance' 'stalling an engine'
The clipped elements of which clipped compounds are composed typically do not exist as independent mora-clipped words (§3.4; It6 & Mester 1992: 26).
77· One of the very few exceptions to the initial rule is aidoru 'idol, star, heartthrob: from Eng. idol. When present as the second element in an assembled compound, aidoru is reduced to its final two moras doru: e.g. caidoru 'child star' from cai1·udot-aidoru; anadoru 'TV announcer heartthrob' from anauNsaa·t·aidoru; baradoru 'TV variety programme heartthrob' from baraetiitaidoru, etc. Since aidoru consistently reduces in this manner, I do not consider forms such as caidoru, anadoru and baradoru to be portmanteaus (§4.3.!.3). 78.
See Footnote 71 on p. 132 for further detail.
Chapter 4. Mmphology, morphophonology and semantics 145
Sekusuaru 'sexual' and hamsumeNto 'harassmenl, for instance, do not exist as the mora·dippings *seku, *hara, or the like79 Moreover, clipped elements may consist of a single heavy syllable: e.g. (219) waapuro, rimokoN, sekohaN, kurokaN, patokt:la, etc. This is illicit in mora-clipping. The homophony brought about by clipping compound elements to a mere two moras can pose problems f()r both Japanese learners of English and English·· speaking learners ofJapanese. The clipped element koN, for example, may be from: (221)
Eng. miss + contest Eng. radio + corztrol Eng. Lolita + complex Eng.jami~v +computer Eng. tour+ cottductor Eng. variable + condenser Eng. general+ wntractor Eng. body + conscious Eng. system + component Eng. summer + concert Eng. Christmas+ company Eng. precast+ concrete
misu+koNtesuto raj io+ko~1:orooru roriita+koNpnreQkusu famirii+koNpyuutaa cuaa+koNdakutaa bariabunHkoNdeNsaa zenerarn+koNtorakutaa bodii+koNsasu Sisutemu+koNpooneNto samaa+koNsaato kurisumasu+koNpanii purekyasuto+koNkuriito
misukoN raj ikoN rorikoN famikoN cuakoN barikoN zenekoN bodikoN sisukoN samakoN kurikoN purekoN
'beauty contest' 'radio-controlled' 'Lolita complex' 'family computer' lour conductor' 'variable condenser' 'general contractor 'body conscious' 'system component' 'swnmer concert' 'Christmas party' 'pre· cast concrete'
Superheavy syllables (1L1, p. 75) contain three moras and we would expect the final mora to be clipped in compounds conforming to the dominant 2f.H 2!1 clipping pattern. Indeed, this is "'"hat we do find in examples such as (220) jiipaN, whose jiiNzu element contains an initial super heavy syllable. 'Ihere are, however, a number of cases where elements with an initial superheavy syllable display irregularity during the compound clipping process. One case in particular is that of super heavy syllables ending in /auN/. Here, we would expect the final iN! mora to be clipped but, as noted by Kuwamoto (1998a) and Kubozono (2001a:67-68, 2002a: 86-87), and illustrated in (222), it is the iu/ mora which is typically elided in such cases. 8 Further, there exist sporadic cases where heavy syllables con·· taining a long vowel are shortened during the 2f.H 2~L clipping process. 81 Some examples are shown in (223).
°
(222)
Eng. sourzdtrack Eng. cormter +stop Eng. Blue A1ountain Eng. Style Council
79· One exception is suto 'hunger strike:
So.
sauNdo·ttoraqku > saJ-.1:ora kaul'.1:aa+sutoQpu > kaNsuto buruu +mauNteN > burumaN sutairu+kauNsiru sutakaN
'strike~
'soundtrack' 'max value (computerese)' 'Blue Mountain (cotiee)' 'The Style Council (band)'
Compare (186) suto '(labour) strike' and (219) lwNsuto
Compare pre-nasal shortening (§3.2.1.2), in particular (106).
Vowel shortening in heavy syllables can also be seen in some verbali1.ed gairaigo. See (199) sutabaru and harnoru.
81.
146 Loanwords in Japanese
(22.3)
Eng. personal computer Eng. birth controi Eng. supercomputer Eng. hello + morning
paasonaru+koNpyuutaa > pasokoN baasu+kol\'torooru basukoN supakoN suupaa+koNpyuutaa haroo+mooniNgu haromoni
~PC'
'birth control' 'supercomputer' 'Hello Morning(TV prog.)'
TI1ere appears to be one type of 2fH2!1 clipped compound which is rarely tolerated, those where /Q/ would occur as the second or fourth mora. \Vhen an element begins in a heavy syllable ending in /Q/, either (i) the /Q/ is elided to produce a 2fl+2fl compound (224); or (ii) the element is clipped to one mora, resulting in a three·mora (225ab) or even a two·mora compound (225c). Given the constraint against word .. final/Q/ in Japanese in general, it should come as no surprise that /Q/ is illicit as a fourth and final mora in 2~t+2fl clipped compounds.82 While a few sporadic examples of second-mora /Qi in clipped gairaigo compounds can be f(mnd (e.g. niQkado, a doublet of (225b), hiQpare from hiQto+pareedo 'hit parade'), its frequent exclusion hom second mora position is curious, given that /Q/ occurs regularly in this position across all vocabulary strata. (224)
Eng. American football amerikaN+fuqtoboom > amefi.tto ~A.merican tootbalf ha.rii.+poQtaa haripota 'Harry Potter' Eng. Harry Potter Eng. net radio neQto+rajio > netoraji '(inter)net radid
(225)
a. b.
Eng. sex +friend seQkusu+ fureNdo sefure 'sex buddy' Eng. nickei cadmium niQkeru+kadomiumu > nikado 'nickel cadmium' Eng. base + ttp
beesu+aQpu
be a
'increase in basic pay'
If the prosodic unit of the foot is taken to consist of two moras (Poser 1990), then, as Kubozono (1999b:40) points out, 'many phonological and morphological structures [in Japanese] ... can be generalized'. The 2fl+2!1 clipping strategy illus .. trated in (219 ..·224) should therefore be viewed as partofa more pervasive proclivity, which also embraces non-gairaigo compound clipping (Shibatani 1990: 254-256), hypocoristic formation (Mester 1990; Poser 1990:81-93), accentuation patterns in compounds (Tsujimum & Davis 1987; Kubozono & Mester 1995; Alderete 1999), sequential voicing (Irwin 2009a), reduplication in mimetics (Poser 1990: 94-95;
82.
There is at least one counter-example (Kuwamoto 1998b: 168) of a dipped Sino-Japanese/
gairaigo hybrid compound(§ 1.2.5): daNtocu 'far and away, runaway' from daNzeN'abs..-:>lute' +
toQpu (from Eng. 'top'). Here, /Q/, the second mora of the second element, has become leu! under the influence of kana: leu/ is-.~,, "'hereas /Q/ is its reduced variant •:1 (§5.1). In the SinoJapanese stratum, but to a far les:;;er extent in the native stratum, /en/ often assimilates to /Q/ before a voiceles:> obstruent: e.g. recu 'rank' but reQsuru 'to rank among: See Vance (1987: 155164) and Irwin (2009c: 74-88).
Chapter 4. Mo.tphology, morphophonology and semantics 147
Hamano 1998: 25-38) and jazz argot (Tateishi 1989; It6, Kitagawa & Mester 1996; Kubozono 2002b).
Although we have seen that bimoraic clipping of elements, resulting in quadri-· moraic 21.1+ 2fl clipped compounds, is the dominant pattern, a few examples (5/455, 1%) of quadrirnoraic 1fH3fl or 3fH 1fl compounds also occur. These are mostly half clippings: e.g. pariigu 'Pacific League (in baseball): from paJ~fiQku + riigu. Monomoraic mora-clippings are illicit (~3.4). With compound clipping, however, monomoraic clipping of an element to only one mora, while at)'-pical, is tolerated. By extension, reduction of compounds to three (226-227) or even two (228) moras maybe found. Only 16%ofthecompounds in my database are oftl1e 2fl+1fl type and a mere 2% of the lf.l+2fl type. Examples of these are shown in (226) and (227), respectively. Meanwhile, just 1% were 1!1+ 1fl, as illustrated in (228). Here, both the examples are obsolescent, as is a further example, (225c) above. Both two- and three-mora clipped compmmds can be either assembled or imported. > > > >
'Mister Donut' 'plastic model' 'tracing paper' 'homepage, website'
(226)
Eng. Mister Donut Eng. plastic model Eng. tracing paper Eng. homepage
(227)
Eng. mail address meeru·t·adoresu > meado 'e-mail address' Eng. tape recorder teepu+rekoodaa > tereko 'tape recorder' Eng. lemon squash remoN+stlkaQsu > resuka 'lemon squash'
(228)
Eng. modem+ girl modaN+gaaru Eng. modern + boy modaN+booi
misutaa+doonacu purasuciQku+moderu toreesiNgu+peepaa hoomu+peeji
misudo puramo torepe homupe
> moga 'flapper' > moho 'dandy'83
Examples such as (226) misudo, torepe, homupe, (227) tereko, (228) moga, mobo provide further grist for Kubozono's (2002a: 94) contention that in Japanese in general 'long vowels tend to be shortened in word--final position: Indeed, (227) meado, tereko show that they may also be shortened in an initial element. Nevertheless, the motivation behind (226) puramo and (227) resuka not following the default 2fl+2fl clipping pattern is obscure. Gairaigo compounds clipped to five or six moras are uncommon. While the former comprise 4% and the latter 1% of my database, all are half clippings. The only full clipping is the 2f1+3fl compound kacusaNdo 'wiener schnitzel sandwich: from ktu:urecu '(pork) cutlet' + saNdoiQCi 'sandwich'.
83. Mobo is a diffi.cult word to gloss in English, as in the West there was no male equivalent to the female flapper of the 1920s. 'Dandy' is probably anachronistic.
148 Loanwords in Japanese
A curious phenomenon which cuts across all these compound clipping patterns, but receives only brief mention in the literature (see, however, Lovins 1973: 140), is what I will term 'mora splitting'. Consider the following examples: (229)
Eng. mass communication Eng. Italian + casual Eng. oJfice computer Eng. staple fibre Eng. differential gear Eng. memory slick Eng. stardust+ review
masu+ko[m.Y..lllnikeesoN itariaN+kafiuiaru olfi.lsn+koNpyuutaa suteepr1rn+~ibaa de[fu]reNsaru+gia memorii+suli]Qku sutaadasuto+rel11Y.i1Ju
> masukoimJ! 'mass media' > itaka@ 'Italian casual (fashion)' > oifuikoN 'ofl'ice computer'
> su[iii 'rayon' > ddJ!i]gia 'differential gear' > memos~ 'removable Hash memory' sutarelhl]
'Stardust Review (pop group)'
Here, the boxed moras in the clipped compounds differ from those in their full forms. In all cases, the full form moras are written with kana digraphs (§5.1): I ) .:::. ~ ~1., {j u ) :;../ '" :r., {fi) 7~ ~,? , (<=1aj' ~-·•'f ) ~-·'""'"" - ? an d {b yu) as \filYU J )', <..a j l ,
Ellipsis
The second major strategy of compound reduction is ellipsis. Here, an entire element is deleted, either the final (230) or the initial (231), while the other remains wholly intact. This strategy is most widespread among imported compounds, although there are a few examples of assembled compounds where ellipsis occurs. These are listed in (232). I ignore here loanwords which may have been borrowed after having undergone ellipsis in the donor language: e.g. supiikaa from Eng.
loudspeaker > speaker. (230) Eng. supermarket Eng. native speaker Eng. front desk
suupaa+maakeQto neetibu+Supiikaa furoNto+desuku
> suupaa > neetibu > furoNto
'supermarket' 'native speaker' 'hotel reception desk'
84. There are also occasional examples of mora splitting occurring in back-clippings (§3.4), including puroJmiru > pun~J.Y] 'profile (on social networks, etc.): from Fr. profile, and ko1'<1li]nyuiti > ko~ 'continuity (in a movie, etc.): from Eng. contimlity. With regard to possible links between certain moraic phenomena and Japanese scripts, see the discussion in Kubozono (1999b: 57).
Chapter 4. Mmphology, morphophonology and semantics 149
Eng. part time (job)
Eng. stainless steel Eng. make--up Eng. notebook Eng. self-service Eng. studless tyre Eng. overcoat
paato+taimu suteNresu+SuCiiru meeku+aQpu nooto+buQku semfu+saabisu sutaQdoresu+ taiya oobaa-t· kooto
paato suteNresu meeku nooto serufu sutaQdoresu oobaa
> > > > > > >
'part time work(er)' 'stainless steel' 'make-up' 'notebook' 'self-service' 'studless tyre' 'overcoat'
(231)
Eng. sewing machine Eng. scre·wdriver Eng. Christmas Eve Du brandpunt 'focus' Eng. greenhouse Eng. tu:wscaster Eng. sandpaper Eng. powder puff Fr. crepe de chine Eng. grandstand play
sooiNgu+misiN sukuryuu+doraibaa kurisurnasu+ibu buraNdo+pil'.rto guriiN+hausu nyuusu ·t·kyasutaa saNdo+peepaa paudaa+pafu kureepu+de+siN guraNdo+sutaNdo+puree
> > >
miSiN doraibaa ibu piNto hausu kyasutaa peepaa pafu deSiN sutaNdopuree
'sewillg machine' 'screwdriver' 'Christmas Eve' 'focus, point' 'greenhouse' 'nev.-scaster' 'sandpaper' 'powder puff' 'crepe de chine' 'playillg to the gallery'
(232)
Eng. American + Du. kojj1e amerikaN+koohii Eng. magic +pen majiQku+peN Eng. televL~ion + talent terebi+tareNto
> >
>
amerikaN majiQku tareNto
'weak coffee' 'marker pen' 'celebrity, TV star'
A much less common form of ellipsis is seen only in infrequently found threeelement compounds. Here, the middle element is deleted: (233)
Eng. background music Eng. soft + ice cream Eng. ballpoint pen Eng. after--sales service Eng. freelatue writer
baQku+guraUN do+myuujiQku sofuto+aisu+ kuriimu boorut-poil'.rto+peN afutaa+seerusu+saabisu furii+raNsu+raitaa
> > > > >
baQkurnyuujiQku sofutokuriirnu boorupeN afutaasaabisu furiiraitaa
'background music' 'ice cream cone' 'biro, ballpoint pen' 'after-· sales service' 'freelance writer'
The ellipsis which has yielded furiiraitaa is especially interesting, given the opacity, even to most English native speakers, of the compound freelance. 85 Perhaps the frequency ofjurii led to an analysis (etymologically correct as it turned out) of raNsu as an independent element. See also (234) furiitaa below. 4·3·1.3
Portmanteau formation
'Ihe final compound reduction strategy, portmanteau formation, is comparatively uncommon in Japanese genemlly. While there are some hybrid portmanteaus, such as gojira 'Godzilla' (from loan gorira 'gorilla' +native kujira 'whale'), there are fewer pure gairaigo examples. Ignoring those borrowed directly, such as sumoQgu 'smog' or taNkini 'tankini' from English, examples are shown in (234).
8;.
Originally, a mediaeval mercena1y.
150
Loanwords in Japanese (23·1) Eng. motor+ Eng. aparlmeut Du. c.alcium +Eng. biscuit Fr. vit1yle + Eng. nyiotl Ger. Energie + Eng. utopia Eng. blizzard + F.JJg. axel Eng. freelance + Ger. •4rbeiter Eng. mst + Eng. hotel Eng. vibrator+ Eng. alarm Eng. yacht + Eng. hotel Eng. Japanese+ Eng. Asian
mootaa+apaat.o karusiumu-1-bisukeq to biniiru-t-nairoN en erugii+ }UUtopia burizaado+akuseru furiiraNsu+arubaitaa resuto-1-hoteru baibureetaa+araamu yoqto+hoteru japaniizu+ajiaN
> >
> >
mopaato karukeQto biniroN enetopia buriseru furiitaa resuteru baiburaamu yoQteru japaniaN
'apartment with garage' 'calcium bismit' 'vinyl on' 'green housing development' 'Blizzard t\.1(e) (manga title)' 'serial part-time worker' 'short-stay hotel' 'vibrating alarm' 'boatel' 'Japanese working in Asia
Kubozono (1999b: 37) notes that a Japanese portmanteau possesses the same number of moras as the second element of the source compound and states further that this appears to be a linguistic universal. 'Ihis would appear to be contradicted by the last five portmanteaus in (234). 4.3.2
Sequential voicing
In Japanese compounds generally, the initial segment of non-initial elements may undergo voicing under certain conditions, an allomorphy known as sequential voicing or rendaku ;1!~ .86 The basic phonological condition under which sequen-tial voicing may occur is that the non-initial element in a compound begin with a voiceless obstruent: (235)
a.
s >z:
nezumi
+
rat
b.
t
>d:
saN
>
calculation
hi
+
fire
tane
nezumizaN 'exponential progression'
>
hidane 'spark'
seed
Motoori (1822) stated that if a voiced obstruent occurs anywhere in the noninitial element, then sequential voicing does not occur. This claim was repeated by Lyman (1894) in English and thus became known as 'Lyman's Law':87 (236)
a.
k > k:
yoko
+
side b.
s > s:
kubi neck
kaze
>
wind
+
suji sinew, tendon
>
yokokaze *yokogaze 'crosswind' kubisuji *kubizuji 'scruff of the neck'
86. See, for example, Otsu (1980), Vance (1979, 1980, 1982, 2005), Ito & Mester (1986), Ito, Mester & Padgett (1995), Rosen (2001), Irwin (2009a). 87. Variations on the Lyman's Law constraint, as well as their history, are discussed in Vance (1980, 2007). Exceptions to the Law are discussed in detail by Martin (1987: 115). Lyman was a foreign advisor (oyatoi gaikok<.A_iin) to the Meiji government: see §2.4.4. Whether he had read Motoori (1822) is undea1:
Chapter 4. Mo.tphology, morphophonology and semantics 151
Providing the non-initial element begins in a voiceless obstruent, and bearing in mind L}'man's Law, sequential voicing ought to occur in any vocabulary stratum (§1.2). In reality; however, the phenomenon cuts across strata to hugely varying degrees. It occurs at a level of around 87% in native vocabulary (Vance 1996: 30-34), with a frequency of approximately 10-20% in Sino-Japanese (Vance 1996:25, App. 2; Irwin 2005: 140···145; cf: (235a) where the second element is Sino·· Japanese), but with only exceptional rarity in the gairaigo stratum. The typical behaviour of gairaigo with respect to sequential voicing is illustrated in (237), where all examples are from English. Sequential voicing fails to be triggered in both imported (237a) and assembled compounds (237b), this despite the second element fulfilling the necessary phonological conditions. (237)
a.
c>c:
b. s >s:
miruku
+ cokoreeto
milk
chocolate
dokutaa
sutoQpu
doctor
+
>
mirukucokoreeto *mi rukujokoreeto 'milk chocolate'
>
dokutaasutoQpu *dokutaazutoQpu 'doctor's orders'
stop
Sequential voicing also fails to be triggered in hybrid compounds where the final element is gairaigo: (238)
f>f: ebi prawn
> ebifurai *ebiburai
+ furai breaded deep-fried food
'scampi, fried shrimp'
The subject of sequential voicing in loanwords is touched on very briefly by Umegaki (1963: 110), Nakagawa (1966: 306), Vance (1987: 141), Kubozono (1999a: 115-118), Ohno (2000: 154) and Tamaoka et al. (2009), the only lengthier treatments being those of Takayama (1.999) and its revised English version, Takayama (2005). In total, these scholars cite only seven gairaigo which undergo sequential voicing: (239)
a.
saNmai three (cards etc.)
b.
iroha
> [01698] saNmaigaruta + karuta 'three-card karuta' card game > [0 1910] irohagaruta +
'iroha karuta'
old name for kana syllabary c.
ama rain
88
+ kaQpa raincoat
> [01732] amagaQpa 'raincoat, rainwear'
Ame 'rain' is an apophonic noun, one of a small group of nouns with variation in their final vowel: a free (or 'exposed') form and a compound (or 'covered') form. Though no longer productive, the most common variation is between final free -e and final compound -a, as found in this example. 88.
152
Loanwords in Japanese d.
tooyu
+
> [01686] tooyugaQpa 'oil paper raincoat'
+ kiseru
> [0 1946] mizugiseru 'hookah'
tung oil, Chinese wood oil e.
mizu
pipe
water
t
kuwae
+
> [0 1777] kuwaegiseru 'pipe'
+ karuka
> kaegaruka 89 'spare ramrod'
holding in one's mouth g.
kae spare
h.
aka red
i.
yam a mountain
j.
iNdo India
ramrod + keQto
blanket + kyaNpu
camping + karee
curry
> [0 1908] akageQto 'country bumpkin' > yarnagyaNpu 90 'camping in the mountains' > iNdogaree 'Indian curry'
Compounds whose second element is karuta, a traditional card game bor-rowed from Por. carla 'card' and a gairaigo with a long pedigree (0 1596), undergo sequential voicing frequently: (a-b) are but two examples and others may be found (e.g. 0 1678 utagaruta 'tanka karuta'; 0 1896 hanagaruta 'flower suit playing cards, hanaji.1da'). Both kaQpa 'raincoat' (+1608 from Por. capa 'raincoat') and kiseru '(smoking) pipe' (0 1612 from Khmer) also have long pedigrees. As well as the examples in (c-f), other compounds are 0 1687 kamigaQpa 'oil paper raincoat= (d)', 0 1839 kazagaQpa 'windproof sleeveless raincoat', 0 1908 natamamegL,eru 'flattubed smoking pipe' and 0 1953 nobegiseru 'metal smoking pipe: Karuka (0 1683 from Por. calcador 'ramrod') too is an early borrowing but I am unaware of any other compounds in which it exhibits sequential voicing. The same lack of any furtl1er sequentially voiced compounds applies equally to keQto, a fore-clipping (193, p. 135) of buraNkeQto (0 +1860 from Eng. blanket). None of the compounds in (239) is common; some are obsolete (e.g. t; h). All are hybrid with the exception of U), a pure gairaigo compound. All are assembled - no imported compounds appear to undergo sequential voicing. It is clear that the early attestation dates of many of the gairaigo elements in (239) are crucial to their having undergone sequential voicing. Part of the language for so long, native speakers have lost any sense of their foreignness and they have been wholly or substantially nativized. For karuta and kaQpa, this sense of
Not listed in NKD, but cited by Takayama (2005: 180) as being attested in the mid-19th century.
89.
90. Neither compound (i) nor (j) is listed in NKD, although both are cited by Nakagawa (1966).
Chapter 4. Mo.tphology, morphophonology and semantics 153
nativization is reinforced by the fact that they are often written in Chinese characters rather than katakana (254, p. 168). Takayama (2005: 178···181) cites a number of other possible reasons to explain the sporadic cases ofloanword sequential vok· ing illustrated in (239). These include phonotactics similar to native vocabulary (but note (i)); the tact that the object represented has 'already lost any connection that would associate [it]'Nith a .f(Heign culture'; and, possibly, stigmatization (note especially (h)). He also points out (opus cit.: 185) that since there are very few minimal pairs distinguished by initial voiced and voiceless obstruents in the native Japanese stratum, 91 sequential voicing can operate there without lexical conflict. This is not the case with the loanword stratum. Here, for example, since kureepu 'crepe' contrasts with gureepu 'grape: voicing would lead to potential misunderstanding. He admits, however, that it is not the case that every gaimigo belongs to such a minimal pair. He also notes that, although the Sino-Japanese stratum exhibits a proportion of initial voiced-voiceless minimal pairs broadly similar to the gaimigo stratum, sequential voicing does occur here, albeit considerably less frequently than in the native Japanese stratum. What is curious about the examples in (239) is the fact that in all cases the sequentially voiced loanword begins with /k/. This is surely mere coincidence: /k/ is the most frequent word·initial phoneme in the Japanese lexicon as a whole and one of the most frequent word-initial phonemes in the gairaigo stratum.
4·4
Semantics
Japanese loam\'ords have been divided into a number of different semantic groupings by various scholars (Arakawa 1931; Ishiwata 1985, inter alia). These include broad-based groupings f(Kused on the user, such as specialist (baioteremetorii 'biotelemetry), business (kuuriNguofu 'cooling-off period') and intelligentsia (peNdiNgu 'pending'), as \·veil as more narrowly-defined groupings based on domain. Ishino (1978), for example, examined the headwords of the specialist dictionary JKS (1978) which divides the terminology it covers into a number of difterent classes. Here, vocabulary related to western sports (golf 99%, baseball 75%) and personal appearance (fashion 97%, beauty 86%) evinced the highest loanword penetration rates, while law (a mere 0.4%) and politics (11 %) showed the least. More commonly, gairaigo are grouped based on perceived motivations for borrowing. The main such groupings include new objects (pisutoru 'pistof, bideo
91. The vast majority of native Japanese words beginning in an obstruent begin in a voiceless obstruent.
154 Loanwords in Japanese
'video'), new ideas or notions (guroobarizedoN 'globalization', manffesuto 'manifesto') and euphemism (moN 'loaD, noomaraizeeSoN 'creating a society in which the physically and mentally disabled can participate' from Eng. normaiization). The most commonly cited wholly linguistic motivation for borrowing is pernicious homophony (Weinreich [ 1953]1963: 57 -58; lshiwata 2001:99-1 00). Homophony is most often encountered in the Sino--Japanese stratum with its more restricted phonotactics. For example, although spelt differently; Sino--Japanese kagaku may mean either 'chemistry' or 'science'. Ambiguity in the spoken language may be avoided by use of their gairaigo equivalents, kemisutorii and saieNsu. Gairaigo often appear to have synonyms in another vocabulary layer, but frequently there exist subtle differences. AQpuru 'apple' and tii 'tea' are rarely f(mnd in compounds, unlike their Sino-Japanese near-synonyms riNgo and ca; wiikueNdo 'weekend' is found with far less frequency than Sino-Japanese ~"'uumacu, while the opposite is true ofepumN 'apron' and native maekake (see Pon2003: 81-211 f<.1r further examples and detailed analysis). During the gairaigo borrowing process, or thereafter, semantic shift may occur. The most common types of shift are narrowing (240) and broadening (241 ). Narrowing is considerably more common than broadening and may sometimes be acute. The language of a rice-based culture, Japanese is well known for hav-ing many words for the staple depending on its state. Raisu, from Eng. rice, refers specifically to boiled polished rice served on a plate (not in the usual bowl) and typically eaten with a spoon or a ft)rk (not chopsticks). (240)
Rus. ikra HKpa 'fish roe' ikura kaNniNgu Eng. bonus boonasu Eng. juice juusu eeru Eng. yell Eng. cunning
(241)
Eng. handle Eng. trump Eng. homepage Eng. jinx Eng. needs
haNdoru toraNpu hoomupeeji jiNkusu niizu
'salmon roe' 'cheating in an examination' 'extra wage payment' 'fruit juice' 'supporting, cheering (a team)'
'handle, handlebar, steering wheel' 'playing cards' 'homepage, website, the internet' '(good or bad) superstition' 'need, demand'
Examples of pejoration (242) and amelioration (243) may also be found, the latter shift being rare. Extension to a metaphorical usage can also be observed, as illustrated in (244). (242)
Eng. commission Eng. claim
(243)
Eng. moody Eng. naive
komiQSON kureemu
muudii naiibu
'bribe, kickback' 'complaint'
'romantic, good atmosphere' 'sensitive, unpretentious'
Chapter 4. Mo.tphology, morphophonology and semantics 155
(244)
Eng. service Fr. piment 'green pepper' Eng. pipe Eng. barcode
saabisu piimaN paipu baakoodo
'freebie, discount' 'airhead, tlighty person' 'personal connections, direct access' 'comb·· over'
Finally; gairaigo .,,,.here the semantic change falls into none of the above cat· egories include: (245)
Eng. feminist Eng. smart Eng. stove Eng. company
feminisuto sumaato sutoobu koNpanii > koNpa
'lady's man' 'slim, slender' 'heater' '(drinking) party'
,-._,
Semantics and morphology converge in a particularly interesting group of assembled compounds usually subsumed under waseieigo in the Japanese tradition (§4.3). Like other assembled compounds, these are composed of two independent gairaigo elements, almost always borrowed from English. Taken together, however, the meaning of the compound deviates from what the semantic outcome of an identical compound would be in the donor language. I shall call these 'semantically remodelled' or 'SR' compounds. As an early illustration of an SR compound, whose first attestation is 0 1930, the two independent English loanwords goo 'go' and sutoQpu 'stop' were compounded to f.ixm the word goosutoQpu. VI/ere an English compound go-stop to exist, a native speaker would probably interpret it as a state meaning something like 'stopping and starting' or 'juddering'. In Japanese, however, the SR compound goosr.1toopu means 'traffic light'. Daulton (2008: 19) points out that it is 'in this realm that linguistic confusion between native speak· ers of English and that of Japanese often peaks' and it goes without saying that SR compounds cause severe problems for foreign language learning. They are perhaps the Napoleons of all faux am is for Japanese learners of English, while provoking bemusement and amusement among English-speaking learners of Japanese. Research is lacking on the motivation behind SR compounding, as well as on its history. A rigorous theoretical morpho-semantic analysis is yet to be published. Pedagogical studies on how much and what kind of interference they induce among foreign language learners are also keenly awaited. In the mean·· time, I offer here what can only be a very brief descriptive overvie\'\'. Some SR compounds contain an element corresponding to the affixes in (206-208) above. These I do not include here, since I prefer to regard such elements as having attained some degree of semantic stability (though sometimes having still been remodelled, viz. sirubaa). The examples I introduce here have elements used, for the most part, uniquely.
156 Loanwords in Japanese
A common motivation for SR compounding is euphemism although, as just discussed, this is a factor in gairaigo semantics in general. Such SR compounds are often related to sex, sexual and bodily functions, or to the sex industry: (246)
cherry+ boy fashion+ health delivery + healttl /Ja/Jy + stop new+ half blue+ day sex +friend stick -r girl soap+ land pitzk + Fr. salon love+ hotel Ger. Onanie +pet pink+ mood out+ sex water92 +girl
ceriibooi faqsoNherusu deribariiherusu bebiisutO\-:lPU nyuuhaafu buruudee seqkusufureNdo suteqkigyaru soopuraNdo piNkusaroN rabuhoteru > rabuho onapeqto piNkumuudo autoseqkusu wootaagyaru
'male virgin' 'massage parlour of dubious repute' 'call-girl service, dial-a-prostitute' 'abortion 'post-operative transsexual maii 'day when one is menstruating 'sex buddy 'female escort' 'bathhouse brothel' 'hand-reliefbar' 'hotel rentl.ng rooms by the hour' 'object of a masturbatory fimtasy' 'sexual overtones' 'extramarital sex' 'bar hostess'
Marketing departments, spin doctors and the media are avid generators of SR compounds: (247)
white+ day golden + hour wide +show
catch+ copy success + look mic + location sigrz +party time + service
play+ guide shutter -r dtance
howaitodee goorudeNawaa waidosoo kyaQcikopii sakusesuruqku maikurokeesoN > maikuroke sai:Npaatii talmusaabisu pureegaido SaQtaacaNSU
'White Day'93 'peak-viewing time' '(low-level, long) TV variety show' 'tagline' 'dark suit worn for job interview' 'on- the-spot broadcasting' 'autograph session 'special offer, sale' '(theatre) ticket agency' 'photoop'
Other semantic fields '"'here SR compounds are often encountered are those of sports (248), the automotive industry (249) and of electronics and telecommu-nications (250). A selection of miscellaneous examples is shown in (251).
92. From ·wootaab~iinesu 'water business: itself a gairaigo euphemism for mizusoobai 'the water trade: which refers to any unstable and fluid line of business, including the sex industry.
93· March 14th, exactly one month after Valentine's Day, when men are supposed to give chocolates to women. Instigated by the Japanese chocolate industry in the late 1970s.
Chapter 4. Mo.tphology, morphophonology and semantics 157 (248)
ground + manners middle + hole point+ getter toss + batting chance + ball guts+ pose out+ course
gurdlL.'Il domanaa midoruhooru poiNtogeQtaa tosubaQtiNgu caNsubooru gaQcupoozu autokoosu
'fair play, sportsmanship' 'par 4' (golf) 'leading scoret' (football) 'pepper drill' (baseball) 'weak return:' (volleyball, tennis, etc.) 'raising one's fist in triumph' 'front nine' (golf)
gasoriNsutaN do doaeNjiN saidobureeki sumooruraito furoNtogarasu supaikutaiya ceNjirebaa
'petrol station, gas station' 'auto open and dose door' 'handbrake, parking brake' 'side light, parking light' 'windscreen, windshield' 'studded tyre' 'gearstick, gearshift'
wookumaN manaamoodo furiidaiyaru teNkii geemuseNtaa kyaQcihoN teresukyaN
'Walkman' 'vibration mode (on mobile phone)' 'freephone, toll-free' 'numeric keypad' 'amusement arcade, video arcade' 'call waiting' 'teletext, dose caption'
makaroniuesutaN pearuQku purasudoraibaa bariafiuii guriiNkaa haroowaaku > harowa baajiNroodo saiNpeN raibuhausu furiisaizu
'spaghetti western' 'his 'n' hers' 'Phillips screwdriver' 'disabled access' 'first class railway carriage' 'government job seekers bureau' 'aisle (in church or wedding chapel)' 'felt-tip pen, marker' 'club with live music' 'one-size-fits-all'
(249)
gasoline +stand door + engine side+ break small + light front+ glass spike+ tyre change + lever (250)
walk+ man manner+ mode free+ dial ten+ key game + centre catch +phone teie +scan (251)
It. macaroni+ western pair+ look plus + (scmv )driver barrier +free green+ car hello t- work virgin + road sign+ perz live+ house free+ size
CHAPTER
.5
Orthography
One of the most salient features of gairaigo orthography is its diachronic instability. Currently, loanwords are typically; though not always, written in a native script used infrequently f()r other vocabulary strata. In other words, although the script in question is multifunctional, its domain of use is skewed heavily towards writing loanwords. That loanwords are thus rendered so conspicuous on a page of printed matter, and in consequence their loanword status perpetually reinforced in the mind of the reader, is unique among the writing systems of the world. This has not always been the case. After considering Japanese scripts in general in §5.1, I examine gairaigo orthography before (§5.2) and after (§5.3) the year 1955, when the Japanese Ministry of Education published important recommendations on loanword scripts and spelling. In the final section of the chapter, §5.4, I look at current use of the Roman alphabet, a script becoming increasingly common in the \'\'riting of gairaigo.
5.1
Japanese scripts
1he modern Japanese language is written, for the most part, in a mixture of three scripts: Chinese characters (known as kanji ~~.oy.~ 'Han letters') and two 'syllabaries: hiragana .:>fJJZ_-4; and katakana l'l· JJZ_-4; . The latter two scripts are known collec· lively as kana ·Jft•.;f... In addition, the Roman alphabet, first introduced hom the mid-16th century by Iberian missionaries (§2.2), has in recent decades come to be used more and more frequently, especially in signage, advertising, pop lyrics, acronyms and computing. Th.e Greek alphabet is employed to a much smaller degree, its usage largely restricted to product names and scientific or mathematical notation. Finally; Arabic numerals, less commonly Roman numerals, and a range of punctuation marks and typographical symbols are also employed. I"igure 5.1 shows a typical Japanese polygraphic newspaper article containing kanji, hiragana and katakana. Loanwords are shaded. In this article all loanwords are written in katakana and all words written in katakana are loans.
160
Loanwords in Japanese
The adoption of kanji in the Japanese archipelago dates from approximately the 5th century CE (Miyake 2003: 5-42; Oshima 2006: 2-9). Kanji could be used, as in China, both for their phonetic and semantic values. The latter usage is relatively transparent. For example, the kanji now written in Japan as ~was used in China to write Early Middle Chinese *wun 'cloud' (modern Bei. yun). After the adoption of kanji by the Japanese, it came to be used to write kumwo, the Old Japanese word for 'cloud' (modern Japanese kumo). 94 Kanji used in this way for their semantic values are known as semantograms.
~~~?~••~~~••m~Y*~tY~-•r~~~? ~7-J ~, jit~lt'~Jv ro~7~'7-J (11sill&dl", 612 j-~~)~--~--~h~~t~~~k. ~Y~77~~ Wf~7h1 S1~:ttc. {J$~:~\J<:::/Yil <. tJ~7t.r~it¢:9 ~ 7"oY::x:. ~ ~tz~td}, ~~~;g~O)~~t.rt-0f<=~~t~~t~~tc. d .:±-f ~$Jltf:Jt.rtE:~ t lt' '5 • l!l-~t5tj',(lfiJ~ 15~ Jv L-tct t1~:t Shclt'~· r tJ~7~?- J ~;U'f&f:iJ>Sd•iJ~ P~* IJ, 20I2&f:fr:jC!JX-f";EtZ~ tc. je!JX9h~i, ~)+i~~t •
r <#1
lt'~Jvt~~-f";Etz~tctlt''5. r ~~~? ~7- J m!!tfi 94&f:~<=rT!>tf:J~h, 9s&f:~~:;;,&;g~~lt'~tc!v•M· -r~~, .®::illl1illf.t?r~••~ ~~iJ~9Twf.l t ~ ~ 0)~<= #lt ,, ••iJ~~Jv ~lt'
tc. ~1<$:~---~i 120~
rlv (#J 1 ~~ 1500~P3) t tlt>:bh
clt'~·
Figure 5.1. Typical Japanese newspaper article containing the three Japanese scripts (Asahi Shinbun web edition, 22 Nov. 2008)
Kanji used for their phonetic values are termed phonograms. Here, a Chinese reading of a kanji came to be used to write all, or part, of a native Japanese word, the meaning of the kanji being irrelevant. For example, Old Japanese kumwo 'cloud' could also be written with the two kanji !:\. and -=€:;, read ku and mwo, together !:\.-'€:; kumwo. The gradual development of abbreviated forms of phonograms - the number of strokes drastically reduced - gave birth to the hiragana and katakana syllabaries from around the 9th century. Although the
94· I use Pulleyblank.'s ( 1991) Early Middle Chinese reconstructions here and further below. The wo in Old Japanese kurnwo 'cloud' indicates, in Yale transcription, a type-A vowel. Vowel phonemes in Old Japanese, and their phonetic values, are a matter of intense academic debate. For a thoroughgoing treatment in English see Lange (1973) or Miyake (1999: 153-170).
Chapter 5. Orthography
two types of kana did not necessarily evolve from the same phonograms, (252) shows some examples where a single phonogram spawned both a hiragan a and a katakana graph. The leftmost column sho>vs the phonograms, in the middle are the modern hiragana and katakana forms, while the rightmost column shows their modern phonemic values. Many kana had earlier variant forms, known as hentaigana JUt·rtz-~;, which have fallen out of use or been officially proscribed. 95 For greater detail on the derivation and history of the kana scripts, see Yamauchi (1972) or Seeley (2000). (252)
~
']
h
/JU
.;,.
$
f.JI •9!
:I.
§
~)
·a
/
/J
u ka yu ro
Table 5.1 shm'\'S the traditional hiragana and katakana series, as reformed in the gendai kanazukai ~W,1\';f,7A<; :It~' 'modern kana usage' government promulgation of 1946. These series are often mistakenly referred to as 'syllabaries: but one kana in fact represents a single mora, not a single syllable (14, p. 75). The leftmost vertical column indicates the initial consonant of the mora; the topmost horizontal row indicates its final vowel (and preceding glide, where applicable). Each cell contains, antidockwise from its top left corner: the hiragana, the katakana, the conservative phonemicization (Figure 3.1), and the contemporary phonemicization employed throughout this book (Figure 3.2). Although the two phonemicizations are frequently identical, the debt owed by the lower conservative phonemicization to the traditional kana series is evident. The g-, z- and d-row kana differ from their voiceless counterparts only through the use of a voicing diacritic , known as dakuten ~ ,(?... 'Ihe b- and p-row kana are composed of the h-row kana with, respectively, an appended dakuten and handakuten #~ .~ ('half dakuten') diacritic 0 • 96 For ease of exposition, the mora consonants /N Q/ are shown together in a separate row not found in traditional kana charts. The fact that they are placed in the a- and i-columns is of no significance. The six shaded cells call for some brief comment. The kana for (wi) and (we), now pronounced /i/ and /e/, date back to the birth of the script but are now quasi -obsolete, confined in the main to a few names.
95· Kana were limited and standardized by law in 1890. See Bailey (1960) and Ishizuka (2007).
96. The cluster /h b p/ is explained by the fuct that modern Japanese non-gairaigo /h/ stems from an earlier /p/. See, for example, Ma1tin (1987: 10-13, 79).
161
162
Loanwords in Japanese Table 5.1. The traditional kana
0
7'
a
'/J' ka k h ka ~ sa s '\]"
.{
i
~
ki ki
"""
si si ci
1._,
to: /;<
n h m
r y w
ta
ti ni ni
:q -t
hi hi
~;l I\
£" ":{
;:_;,
7
u ku ku
e ke ke se 1! se l te 7 te
su ;;<_ su :::> cu 'Y tu N'J nu .tl ne )( nu ~ ne ~ fu ""' he hu ""' he mu ~ me mu j. me n re ru ru v re
0
L. :;J
.:f '.)
t 10)
J ~~
*~ .:c 6
ko ko so so to to no no ho ho mo mo ro ro
1._,-\t>
Y-'r
'S-'It> =f-->r ~C:::-\t>
.::::->r LJ--\t>
1:::->r 'JJ-\t> ~-'r
I? "'t>
-\0 "\? b '7
g z
d b
p N/ lv Q y
N
?
N
'Y
Q
The kana for (wo) are confined in modern usage to writing the direct object particle o, typically pronounced /o/, as well as, once again, to a small number of names. There have never been kana for (yi) or (wu), although the katakana r and T appeared briefly in an 1873 government-sanctioned elementary school textbook (Mabuchi 1993: 17-24). The kana for (ye) are now pronounced /e/, but here the kana for (we) have been used historically in their stead and the katakana ~ appeared in the elementary school textbook just mentioned. One graph, not held to be kana, is omitted from Table 5.1. This is the length mark,
Chapter 5. Orthography 163
known in Japanese as the clu5onpu ·*i·4\L onbiki ·'i% ~~ ~ or, more colloquially, the nobashibo }iHf L:l*. It is t)'"Pically employed in katakana only (very occasional use in hiragana may be found) and is the conventional method in the modern language for indicating the second mora of a long vowel (see, however, (270-271) below).It is written- when writing horizontally, and I when writing vertically. 97 Kana are unicase, i.e. they have no upper/lower case or majuscule/ minuscule variants. In the traditional system, reduced variants of some kana are used to indicate a glide. The ya-, yu- and yo-columns in Table 5.1 exemplify this. While J}. ~ represents the single mora /mya/, J).J?? represents the two moras /mi -ya/. Before the 1946 promulgation of the gendai kanazukai reduced variants were rarely encountered. The huge influx of gairaigo in the modern period has led to the breakdown of the conservative phonemic system (§3.1). Given the consensus that kana should be an accurate representation of pronunciation, it has also meant the traditional kana in Table 5.1 are no longer adequate. The contemporary kana series is shown in Table 5.2. Here, the mora-initial consonants in the leftmost vertical column are the contemporary phonemes in Figure 3.2, while the morafinal vowels (and preceding glides, where applicable) are shown in the topmost horizontal row. In each cell, below its contemporary phonemicization, appears a katakana graph, digraph or, in some cases, trigraph. Although hiragana variants do exist, their usage is highly marked for gairaigo and, in any case, their form may be readily inferred. A shaded grey cell indicates the katakana appear also in the traditional series (Table 5.1). A blacked out cell, meanwhile, indicates the katakana cannot (e.g. (wwa)), or do not (e.g. (jyu)), exist. At the lowest level of each cell may also be found one or more letters, some with a superscript. These refer to the usage recommendations found in the government documents Monbush6 (1955) and Bunkacho (1991) to be examined in detail in §5.3. Approximately 38% (96/251) of the kana. in Table 5.2 have not been officially recommended in either document (e.g. (tya)), many being used to write gaikokugo rather than gairaigo. 98 Table 5.2 is not the kana chart taught in Japanese schools. In spite of government recommendations (§5.2, §5.3)
'Ihe second mora of a long vowel may aL~o be indicated by an outsi7.e stand-alone tilde (SMS messaging) and in manga. Kana iteration marks (odoriji BWi kurikaeshifugo ~I):![ L..it-%), such as the hiragana :. or and the katakana . . or .._,,have mostly fallen out of general use and are now found only in a few names.
97·
~.especially in texting
98.
I)*·
A$ transcriptional conventions differ according to foreign language, in some cases a gtli-
knkugo value may not necessarHy correspond to that found in 'Iable 5.2.
164 Loanwords in Japanese
necessitating the extra rows and columns in Table 5.2, Table 5.1 remains the pedagogical standard. Five columns not found in Table 5.1 appear in Table 5.2: the ye-, wa-, wi-, weand wo-glides. Two types of wa-glide appear: those written with a reduced 7 (a) and those written with a reduced r; (wa). While those digraphs appearing with a reduced 7 (a) may also be written with a reduced r; (wa) (e.g. (kwa) may be 7 7 or 7 r; ), the opposite is not the case. If a digraph with a wa-glide written r; (wa) is written instead with 7 (a), it will represent a different mora: e.g. 7 r; is (fwa), but 7 7 is (fa).
Table 5.2. The contemporary katakana series
pa
pi
}~
\:': B' M
M
ba
bu 7"
B' M
M
pe
po
~
~
B' M
bi 1!:
}~
ta 51
pu 7"
B' M
B' M
B' M
be
bo
""'
~
B' M
tu
te
to
7-r
l-'7
=;-
}-
B' MF B' M
pyu l:':::z.
B' M
bya 1!:-v
B' M
ti M
pya 1:':-y
pwa
pwi
pwe
l:'::s
7"7
7"-r
7":r.
B'
pwo 7":t
B'
M
bye
byo
bwa
bwi
bwe
bwo
l!:::z.
l!::r.
l!::s
'-;/7
'-;/-{
•f:r.
/';t
B'
B'
M
tyu
tye
tyo
twa
twi
twe
two
7-::z.
7-:r.
7-a
l-7
l--r
!-:r.
}-;,)-
B'
B' M
pyo
byu
B' M
tya 'T-v
pye l:'::r.
B'
da
di
du
de
do
dya
dyu
dye
dyo
dwa
dwi
dwe
dwo
~
7-r
F7
7
}"
7-v
7::z.
7:r.
TE
1"7
F-r
F:r.
F:t
kwi 7-r
kwe
kwo
7:r.
7:t
M
B' MF B, M
ka
ki
tJ
~ M
ga
iJ' M
ca ':!7
ku 7
B' M
B' M
ko
kya
kyu
kye
kyo
kwa
::J
~"\'
~::z.
~I.
~E
77
ge
::¥
-1
';f
·:;-{ s>
B' M
go :i B' M
ce '/'I.
'
B' M
gya ::¥-v
B' M
co 'Y :t
B' M ' B' M
M
B' M
B' M
gu
cu 'Y
B'
ke
gi
ci
M
7
B' M
B' M
B'
B'
gyu ::¥::z.
B' M
'
B
gye ::¥:r.
B'
B' M B,
M
gyo ::¥:s M
'
gwa 7'7
B' M s>
cya
cyu
eye
cyo
cwa
'/'"''
·:; ::z.
·:;I.
'Ya
'/''7
ca
Ci
cu
ce
co
ewa
'f-
7-::z.
7-:r.
7-:s
7-'7
B' M
B' M
B' M
7'-r
'
B
cwi
s>
s>
gwe 7':r.
gwo ij';t
B
'
B
ewe
CWO
·:; '7 -{ ·:; '7
I.
'
'Y '7 ;t
B'
7--v M
s> gwi
cwi
ewe
CWO
7-::z.-r 7-::z.:r. 7-::z.:t
B'
ja
ji
ju
je
jo
jwa
jwi
jwe
jwo
~"\'
~
~::z.
~I.
~E
~'7
~::>.-{
~::I. I.
~::z.;,f
M
B' M B' M
B' M
B' M
B'
(Continued)
Chapter 5. Orthography 165 Table 5.2. (Continued)
77
7-( 81
81 M
81 M
M
B'
81
81 M
M
va
vi
vu
~
w
~
~
~
~
~
~
r'}y
r'}-1'
r'}
r'}.r.
>;l'>t
>;~'""
>;i'.:z.
.;i'-1 .r.
>;l'a
>;i'r;
>;i'r; _,
~
B'
B'
82
82 M
82 M
82 M
82 M
M
~
>;i'r; .r. >;i'?>t
sa
si
su
ro
~
~
~
~
~
~
~
~
-lj'
A-1
A
'J
A"\'
A.:z.
A-1'I
Aa
A7
A"J-1'
AI
A>!'
81
81 M
81 M
B' M
81
M
=
za
zi
zu
ze
-If
A -1
A
-t!
§a
si
su
se
So
swa
Swi
Swe
swo
~"\'
~
~.:z.
~I.
~a
~'J
~'7-1
~?I.
~?>!'
B' M
hi
hu
/\
1:::
:;j;ry
mi
¥ ;(a
~:7
~
~
~
A?-1
A.r.
A>!'
ho
hye
hyo
hwa
*
l:::.r.
t:::a
:;J;r;
*"'
8 1 M' 8 1 M
81 M
81
hwi
me
mo
mya
myu
mye
myo
mwa
mwi
mwe
mwo
b.
:;1.
'C
~"\'
~.:z.
~I.
~a
b.7
b.-{
b..r.
b.>!'
M' 8 1
81
ne
00
~
~
~
•
~
~
~
~
;f.
/
.::.""
.::..:z.
.::..r.
.::.a
Y.7
Y.-1
Y..r.
Y.>t
8 1 M' 8 1 M
81 M
81 M
81 M
8 1 M' 8 1 M
81 M
81 M
nu
ni M
""'
• A-1.r.
mu
81 M
M
he M
81
M
~
A.:z.
81
81 M
81 M
ha
• A"'r
81
81 M
81 M
81 M
81 M
M
w '.!
81
ri
ru
re
ro
rya
ryu
rye
ryo
rwa
rwi
rwe
rwo
l)
tv
v
p
l) "'r
l) .:z.
l) .r.
l) a
!117
111-1
Jv.r.
Jv>!'
ya
yi
yu
ye
-v
-(-{
::I.
-(.r.
yo 3
M
wa '7
81
81
M" 8 2 M
wi
wu
we
wo
'7-1
'7?
?.r.
'7>1'
82
M
N
M
82 M
82
Q ''j
/
M
81
No original graphs appear in Table 5.2, although the v-row katakana make use of dakuten on tJ (u). The digraphs tJ ...(, tJ .:r. and tJ ;t are officially recommended for /wi/, /we/ and /wo/ (compare the katakana (wi) Jf., (we) J: and (wo) 7 in Table 5.1). As in the traditional series, a reduced katakana frequently indicates a glide. In the contemporary series, however, it may also be employed to
166 Loanwords in Japanese
indicate moraic unity with the preceding full-size graph. This is exemplified in the katakana for (ce), ~'/ .J:.. Here, the reduced .::c. does not indicate a glide, rather that .:r:. should be read together with '/,replacing its final vowel. In modern Japanese the domain of use of the three major scripts is, by convention, as in Figure 5.2. Although most native speakers follow these conventions in their writing practice, considerable licence exists, particularly in advertising, brand and corporate names, poetry, manga, texting (SMS messages) and e-mails. See NINJAL (1983a), Smith & Schmidt (1996) and Tranter (2008) for issues of variation. In Figure 5.2, the term 'sino lingual' refers to countries or territories where a Chinese topolect is an official language (PRC, Hong Kong, Macao, Taiwan, etc.). As is readily apparent from Figure 5.2, script usage in modern Japanese is highly complex, with katakana displaying the widest functional range of the three scripts. 99 Script functions can be broken down into five major categories: (a) the writing of names, both foreign and native; (b) the writing ofloanwords, excluding names; (c) the \'\'riting of non-loanwords, excluding names; (d) indicating kanji readings for pedagogical purposes; (e) other functions. The order (a)···(e) is not intended to be indicative of any ranking, but is employed for expository purposes. It is categories (a) and (b) with which I will concern myself in this chapter. With category (a), names, orthographic practice largely favours kanji or katakana, with hiragana usage f(mnd only occasionally. The names of Japanese nationals (al), as well as nationals of sinolingual countries (a2), are typically written in kanji. Meanwhile, the names of citizens of non-sinolingual countries are written in katakana (a6), although kanji are also found for South and North Korean nationals. Recent NHK broadcasting guidelines (NHK 2006: 40) echo these conventions (§2.5). Japanese place names are written overwhelmingly in kanji, with rare instances of hiragana or katakana (a3). Further detail here need not concern us. Chinese place names too are written almost exclusively in kanji (a4), except where a character is non-standard in Japan. With the exception of Korean place names, where kanji spellings are still found, non··sinolingual for·· eign place names are now written in katakana (a7), although formerly this was often not the case (267). The abbreviations for the more significant foreign countries, many borrowed from Chinese, are exceptions (aS). These are typically still
99· In addition, katakana is, or has been, employed as a script by languages other than Japanese, e.g. Ryukyuan, Ainu and Taiwanese. In such cases, graphs, digraphs and trigntphs different from or supplementary to those in Table 5.2 may be found.
Chapter 5. Orthography
KANJI al
names of Japanese nationals (occasional hiragana or katakana)
a2
aS
names of sinolingual country nationals (occasional hiragana or katakana where kanji is non-standard in Japan) Japanese place names (occasional hiragana or katakana) place names located in sinolingual countries (occasional hiragana or katakana where kanji is non -standard in Japan) and the Korean peninsula (katakana also found) abbreviations of major countries, regardless of sinolingual status
b1
Chinese gairaigv (also katakana)
c1
nouns belonging to native Japanese and Sino- Japanese strata (occasional hiragana where kanji is unknown, difficult or ambiguous) the stems of verbs and adjectives belonging to the native Japanese and Sino-Japanese strata(occasional hiragana where kanji is unknown, difficult or ambiguous)
a3
a4
c2
HIRAGANA c3 c4 cS
verbal and adjectival inflections grammatical elements (particles, conjunctions, auxiliaries, etc.) words belonging to the mimetic stratum (also katakana)
dl
Japanese readings (semantograms, kun'yomi If~ ~J;.) of kanji when listed in dictionaries and educational materials
KATAKANA a6 a7
names of non-sinolingual country nationals (kanji also found for South and North Korean nationals) place names located in non-sinolingual countries (kanji also found for Korean place names)
b2 b3
all gairaigo other than Chinese (very occasional kanji or hiragana) gaikokugo (4. p.lO)
cS c6
words belonging to the mimetic stratum (also hiragana) names of flora and fauna regardless of stratum (also kanji and hiragana)
d2
Chinese readings (phonograms, onyomi if~J;.) of kanji when listed in dictionaries and educational materials
el e2 e3 e4
'foreigner talk: broken Japanese, infant babbling emphasis telegrams and telexes (now largely obsolete) marked (substandard, non-standard, slang, stigmatized) meaning or usage
'igure 5.2. Conventional script usage in modern Japanese
vritten in kanji and derive from former orthographic practice: bei *'USA: ei ~ UK', goo 'Australia',focu {L 'France: ro '4 'Russia: etc.
*
168 Loanwords in Japanese
·when names are ignored, conventional script usage for gairaigo is marked by a division between Chinese (bl) and non--Chinese (b2). Perhaps more than 9996 of the latter are \\Titten in katakana, and in katakana only; 100 as illustrated in (253). A legacy of former orthographic practice (§5.2), some of the very few which may also be written in kanji are listed in (254). Which of the two scripts is employed is governed by individual preference or editorial practice. Chinese gairaigo, on the other hand, are most often written in kanji, though katakana may also be found, as illustrated in (255). Both Chinese and non-Chinese gairaigo are found with only extreme rarity in hiragana, with examples generally confined to the oldest loans. For example, (254---255) tabako, teNpura, karuta, kaN, raameN may occasionally be written k: ~;f C., T !uli- 6, 'b~-'f:J tc::, t;~ fu, ~) "JD"thfu. Mixed orthography; though infrequent, may also be found: e.g. (254) teNpura JZ:.~G, (255) raameN -;;~~. Ruby, kanji glosses appended above or to the side of a gairaigo, may be used for neologisms and for loans deimed ~kely to be unfamiliar to the reader. Thus, (253) 'renewal' may be \\Tit!en 1J -=:. ::1.- ·T ;I:, or the position of gairaigo and gloss can be V-::"- """"/IV reversed to give AF~ ~rr . (253)
Eng. renewal
(254)
Por. tabaco Por. capa Por. tempera 'seasoning' Por. carta Du. koffie
rinyuuaru Du.jodium yojiumu Fr. onomatopee onomatope Ger. Spur 'track, trace' supuuru
Du.glas Du.kan Eng. page Eng. catalogue (255)
tabako kaQpa teNpura karuta koohii garasu kaN peeji katarogu
1) . --:.:. .:1. ...... ·-.;;- ; )/ 'renewal' 'iodine' 3:?'71.:. 'onomatopoeia' ;;d- /''?' l' -~ 'ski tracks' 7°·----JL'
, 'cigarette, tobacco :)< .7\:::I }J•;;J\ 'raincoat'
Jf.Jf.-11( ~-~
'tempura'
7->-'7°7 7~~HV7~~1\
'playing cards' 'coffee' '(pane oD glass' 'can, tin' 'page' 'catalogue'
iJ Jv:Sz
lttl $/.g,~~
:1-1::- ~!f.!~
7J7A
'ii}\j-f-
iJ/
·ffl
-"\--····-::.-'
~
:iJ5u7' ~-~
7--'f·----}\/ Bei. chdofan ;K;Ji-fR caahaN 'fried lice' ;K;Ji'l ~---- ~-;- --v :,../ jf}1);: Bei. majiang A*h~ maajaN 'mahjong' Bei. lamian #lil1 raameN 'noodle soup' #Jii!i:$;-~, etc. .;:;:;~;<./
100. It is, of course, possible to assign kanji to any loanword should one desire and examples of this practice can readily be found, especially in word play and advertising. But this is neither necessary, nor is it the modern convemion.
Chapter 5. Orthography 169
5.2
Loanword orthography until1955
Orthographic practice during the Iberian borrowing phase (mid-16th century to mid--17th century) was characterized by extreme instability. This reflects not only the lack of any orthographic traditions for loanwords, but also variation in pronunciation. As related in §2.2. Jesuit missionaries introduced Portuguese, and to a lesser extent Latin and Spanish, words to their followers. They had early on abandoned calquing Catholic terminology due to misinterpretation by Japanese speakers, a major source of which was the kanji in which the calques were written. While a few of the more common religious borrowings, and a greater proportion of Iberian mercantile borrm-\oings, were \-\-Titten in kanji (259-261), for the most part himgana, rather than katakana is found (Doi 1933: 15). Jesuit missionaries were the first to use in print (Society ofJesus 1598) the diacritic 0 , the handakuten, placed on the h-row kana to indicate /p/ (Table 5.1). They did not invent it, however: according to Bailey (1960: 301), its first recorded use was in an illustrated scroll of 1536. Previously; the kana i \,for example, had been ambiguous and could represent either /ha/ or /pal (or even /ba/). With the diacritic, ;-~is unambiguously /pal. The missionary literature of this period also includes the first Japanese texts, and gairaigo, written in the Roman alphabet. While long /uu/ was consistently written in standard size, the long vowels /aa ii ee/ f(mnd in the gairaigo of the period were transcribed differently according to kana script (Doi 1933: 15-16; Umegaki 1963: 125). This is illustrated in (256-257). Although the relevant 0-row kana was used for the second mora of the long vowel in both hiragcma and the less commonly employed katakana, it was written standard size in the former (256) but reduced size in the latter (257). The attestation dates in (256) and elsewhere in this chapter are those of the spelling, not the word itself; those in (257) are from Doi (1933). The phonemicizations in (256---261) are those of the modern pronunciation. 'priest, father' ~
(256)
[0 1593) [0 1600)
(257)
[1600] [1600] [1600)
Por. missa
Por.filho Lat. pater
miisa hiiriyo paateru
'mass' 'son' 'priest, father'
.::...r-lf "l::: ..( ~) ::1
'""7-iv
Both kana (ou) and {au) are found for long /oo/. This confusion was the result of a series of monophthongizations, known in the Japanese tradition as chocm -l-tf and occurring from the turn of the 14th century, which affected Nul vowel sequences (Toyama 1972: 209-222; Martin 1987: 44; Irwin & Narrog
170
Loanwords in Japanese
forthcoming). While (ou) unambiguously reflected a pronunciation *[o:], it is unclear whether (au) represented *[o:] or*[:>:], a sound no longer found in modern standard Japanese. (258) [0 1592] Por. cathoiica katoorika 'Catholic' 7J l- rJ 1) 7J [0 1666] Por. veludo biroodo 'velvet' tS: f)? 2 Wben kanji were used to write loanwords in the Iberian period, they were employed as semantograms (259), as phonograms (260), or as both (261). Many Iberian mercantile loans may still be found written in kanji, as well as in katakana: see the examples in (254). In contrast, none of the few religious kanji spellings dating from this period are still in use today outside specialized circles. 'tobacco' (259) [•1605] *'- 'smoke'+ J$- 'grass'=> fw..:f. tabako [0 1608] *-*~ 'swan'+#). 'woollen cloth'=> f<J~~#J. biroodo 'velvet' 101 (260) [0 1568] A iru + 7i maN=> />-.Ji irumaN 'lay brother' [161sj1°2 Jt, maru + Jiil Ci + '{g ru => Jt, :fu WI maruCiru 'martyr' (261) [0 1610] f.!' bac...- 'together'+ fi.: teN 'heaven'+:)#, reN join'=> f.!'fi.::lt batereN r·}1639] i!t ki 'luck'+ ftl ri 'benefit'+ J:. Si 'support'+ Jt taN=> w;;
'priest' 'Catholic'103
In a reversal trom the Iberian period, orthographic practice during the Dutch borrowing period (mid-17th to mid-19th century; §2.3) tended towards writing gairaigo in katakana rather than hiragana (Ishiwata 2001: 241). It also exhibited a strong tendency towards using kanji. The majority of kanji spellings from the Dutch period were written using phonograms (262). Examples of kanji used as semantogramsin this period include those in (263). Manyofthesewouldbe better described as calques: biiru ~51W, for example, could also be read as Sino--Japanese bakusu. Many spellings were unstable and the examples below exhibit variation. A few kanji spellings established in the Dutch borrowing phase are still in use today (25L1), but most Dutch gairaigo, especially medical and scientific vocabulary (Table 2.5), are now spelt in katakana. (262) [•1816] ~ ki + ~~~ na :::::>~~~~kina 'cinchona tree' (0 1833) Jli!. a+ at te +~if. ru ====> .llt at~. aateru 'ether' [0 +1837) ~ mo + 1lt ru + l:t hi+ tM ne ::::> *-11Htft!. moruhine 'morphine'
101.
'Ihe allusion to 'swan' is due to the perceived lustre of white velvet (NKD).
102.
Attestation date is from Kobayashi (2009:9).
Many other spellings are recorded for kirisitaN, including politically motivated ones postdating the prohibition of Christianity by the shogunate. These include Jl!.-1'~ ¥.Y/i,i 'lit. devilbenefit-unto-end' r:'1656], Jt!ft~li 'lit. devil-benefit-(phonogram)-end' r:'1676] and --iJJJL;It 'lit. cut-death-(phonogram)' r:'1698].
103.
Chapter 5. Orthography
(263)
[+1811] ~~ 'barley, grain' + JliJ 'alcohol' ==> it Jm biiru 'beer' [+1851] ''iW 'pour' + f,ij- 'tube' ==> P®J 1m poNpu 'pump' 'pistof [0 1871) 'short' + t:t 'gun' ====> ~Jr~JE pisutoru
m
Some phonograms in (262) appear to bear no resemblance to modern Japanese readings, e.g. the use of ;it, 'infant' (or :Ef· 'ear') for the mora ru. This is due to the fact that many scholars of the period were more familiar with W-in readings (§2.5), consulting Chinese translations of Dutch scientific texts before embarking on their Japanese ones (Saito 1967: 1). 'Ihe to-in readings of both JG 'infant' and :EF 'ear' are ru. The use of a choonpu ,____ to indicate long vowels in gairaigo (e.g. ::::1------ /koo/) first appeared in the early 18thcentury(Arai [1715] 1968), while the use of reduced kana (e.g. 7 -1 /wit) can be found from around the close of the Dutch borrowing phase, i.e. the first half of the l9th century (Umegaki 1963: l26). Well into the Western borrowing period, gairaigo continued to be written in kanji, usually with appended ruby. Research by Hashimoto (2010: 141) has shown that, in the 1910s, 53% of gaimigo appearing in Asahi Shinbun editorials were printed in kanji. By the 1930s, this figure had fallen to 8%, and from the 1950s onwards no examples appear at all. In an analysis of the novel Sanshiro by Natsume Si'.iseki "%.. fl {'tk/Ei·, published in 1908, Kobayashi (2009: 25-56) f(mnd that approximately 75% of all gairaigo tokens were written in kanji. These were typically semantograms (264), but phonograms also occurred (265). As indicated in (266), ingenious combinations of both semantograms and phonograms are also found in this period. With the exception of (265) romaN, from Fr. roman 'novel, courtly romance: all the examples in (264-266) are borrowings from English. (264)
5~ ~
(265)
1~ .~
(266)
'Western' + tt 'stick, cane' .-.-.-.-> 11~~ suteQki 'walking stick, cane' 'phosphor' + -t 'sun, inch, a little' ==> #f t maQci 'match' ro +-~maN.-.-.-.-> m_~ romaN 'novel, adventurous spirit' ba + )Jt kecu ==> .~}fl. bakecu 'bucket'
[0 1897] 1* ku 'together'+*' raku 'enjoy'+%~ bu 'club' = } {~k*-llP kurabu 'club' [0 1928] llli. kata 'model'+ :$!f; roku 'record' ~ !Jll.:flt katarob'll 'catalogue'
vVhen not written in kanji, gairaigo continued to be written predominantly in katakana, as they had been since the mid-17th century, and as they are today. For some decades in the early 20th century, though, they were often written in himgana., especially in literary works (Umegaki 1963: 132). Since, until the mid--20th century, katakana was often employed for functions c3---c5 in Figure 5.2, i.e. where hiragana is fc)lmd today, gairaigo were sometimes underlined to distinguish them from non-loan vocabulary.
171
172 Loanwords in Japanese
Highly unstable spelling also continued into the early Western phase. Umegaki (opus cit.) and Ishiwata (2001: 151), for example, both quote a survey by Kojiro Tanesuke 1~ ·R. ~-1E which cites 29 different kana spellings f()r Goethe recorded in a 56-year period between 1872 and 1928. These include ~j';t ::¥3- ''!, ·:7'i· :I-·'?, :::J.:r. T and;::\:!:' :1. --T. Another example of extreme orthographic instability is the word for 'whisk(e)y': NKD records '7 4-.A:t--, 4- .A~t--, 4-/,~t-, ,,_, ··( ,A:t-, 'J .--( ;;\~~-, 'J4<:Z·'T and 'J :<.'T, as well as the stable modern form 'J-( ;;\:'}·----. According to Umegaki (1963: 98), the first use of the v-row kana is found in the English dictionary edited by Fukuzawa (1860): e.g. 74'1Jv (fivuru) 'fever' and T'f7 .::;. :f > (vomiciN) 'vomiting'. The practice was thus reserved initially for indicating the pronunciation ofgaikokugo, rather thangairaigo. Bothf- and c-row katakana also came into use in the early part of the Western borrowing period.
·c,
Unlike current practice (Figure 5.2), names of non--sinolingual countries (267), as well as of non-sinolingual nationals (268), were typically written in kanji throughout the Iberian, the Dutch and the earlier part of the Western borrowing phase. Katakana spellings did not begin to outnumber kanji spellings until the second quarter of the 20th century (Fujimoto 1993; Hida 1998; Ide 2005 ). Kanji spellings were often borrowed directly from Chinese. The examples in (267--268) are taken from AGJHI (1979), but other spellings can be found. For each example, the main historical kanji spelling is on the left along with its pronunciation- other pronunciations were possible. This is follm\'ed by the most commonly found mod-ern katakana transcription and its phonemicization in some cases this modern pronunciation is different from the major historical one. (267)
i'IH~:f,f:.
u;.< PfflP(D,
daburiN Ciri nyuuyoruku teheraN jagatara
*
reeniN
~~=f 5f~
il1l'ff :;~·rt
(268) j1J
.f#:M]~.ft ~UVf· :~ 'j i:J>~ ,~'*~
:::. ~- •····· :::T •···· 7 ~r"""'7'Y
v
:;:; -'f ill :)<
]/-.:::: /'
/:i:J /' bisumaruku ·!::': ;:z '""< Jl, )7 seekusupia :/ :L -7 .'7.1::~7 mahomeQto ·'f:J\/7 F rLNkoru."
Jt
).;f'.-Y' 11 ~ / .7- / ... · '
1)
daburiN ciri nyuuyooku teheraN jakaruta
'Dublin:' 'Chile' 'New York' 'Tehrati 'Jakarta'
reenLN riNkaN
'(Vladimir Ilykh) Lenin' '(Abraham) Lincoln:' bisumaruku '(Otto von) Bismarck' seekusupia '(William) Shakespeare' mohaNmado 'Muhammad (ibnl\bdullah)'
The Japanese government began issuing guidelines on gairaigo orthography in 1902, although recommendations on standardization (Monbush6 1902) were initially limited to names. Further guidelines were announced later (RKC
Chapter 5. Orthography 173
1926), but then rescinded -the focus of Japanese government language policy in the first half of the 20th century was rather on the reform of kana usage in general, along with setting limits on kanji use (Twine 1988: 453--454; Gottlieb 1995: 5L1-85). It was two major reports published in the second half of the century, Monbush6 (1955) and Bunkacho (1991), that had the most impact on how loanwords are written today. These I discuss in detail in the following section.
5-3 5.3.1
Loanword orthography from 1955 Orthographical recommendations I: Monbush6
Monbush6 (1955) was the result of two years of deliberation, from 1952 to 1954, by the Technical Terms and Orthography Joint Subcommittee (Jutsugo Hyoki Giido Bukai) of the Second National Language Council (Kokugo Shingikai). 104 Given centuries of instability, it was probably inevitable that no overall agreement on a standardized loanword orthography was reached (opus cit.: 1) and that the fruits of the Subcommittee's deliberations presented to the Council in March 1954 were contained in a report rather than a formal recommendation. Before listing 19 'General Rules' for loanword orthography, the preface to the report (opus cit.: 3-4) makes clear some of the reasons for the lack of consensus achieved on standardization. The Subcommittee considered there to be three types
ofgairaigo: i.
Those with a long history, which have 'fused' with native Japanese and which the average speaker does not feel to be gairaigo. Examples cited include tabako 'cigarette, tobacco' and kaopa 'raincoat: ii. Those which are probably felt to be gaikokugo rather than gairaigo. Examples cited include oosoritii 'authority' andjiaNse 'fiance(e)'. iii. 'Ihe remainder, i.e. those which are in widespread use and are felt to be gairaigo. Examples cited include oobaa 'overcoat' and ra_iio 'radio'.
104.
The Technical 'Ierms Subcommittee had 13 members and was chaired by Dr. Satta Kotoji author of books on psychology and phonetics and Professor at Tokyo University of the A1ts. The Orthography Subcommittee had 15 members and was chaired by Dr. Hoshina Koichi !¥:#~---,author of numerous books on the Japanese language and Professor Emeritus at Tokyo Bunrika University (later to become pa1t of Tokyo University of Education). The Second National Language Council's 62 members were chaired by Dt: 'loki Zenmaro l:l~t* ~.poet, head of Tokyo Hibiya Metropolitan Library, and active in both the romani1.ation and Esperanto movements. -~l'B!!$*._,
174
Loanwords in Japanese
It is noted that, although for (i) spellings have been largely fixed, for (iii) nativization has meant spellings are frequently divorced from donor spelling or pronunciation, while for (ii) spellings close to donor spelling or pronunciation are still being sounded out. In these references to 'donor spelling or pronunciation' (note the order), the power of donor orthography is made dear (§1.1, §3.2). The first two General Rules concern themselves with overarching issues. Rule 1 states that gairaigo should be written in katakana and specifically recommends the 113 (di)graphs coded Min Table 5.2. While ::1. (dyu) is recommended, its voiceless counterpart ::;- ::~. (tyu) is not. A further 12 digraphs, including the bulk of the c-row but, oddly, some traditional kana (Table 5.1) such as:::. 2:3 (nyo) and C: --'f {hya), appear in parentheses. TI1ese are coded as M~ in Table 5.2. Rule 2 urges following accepted spellings where these exist (e.g. -7-:t-- keeki 'cake'), although this is often contradicted by subsequent rules recommending standardization. No attempt is made at defining what an 'accepted spelling' might actually be. Rules 3 to 5 deal with the mora consonants IN! and /Q/. Rule 3 states that the mora nasal /N/ be written Y. Contemporary spellings with L. can be found: e.g. ';/ .Z.,.;.)t::/1) L simupojiumu 'symposium'. Rule 4 recommends the mora obstruent /Q/ be written reduced, as Before the 1946 gendai kanazukai (§5.1), standard size'/ was common, not only ingairaigo spelling but across all vocabulary strata. Rule 5 states that neither of the mora consonants should be inserted on the basis of dmible letters found in donor word spelling: e.g. :.1 ~ ::i. :::/;- komyunike not ::1 ;::.t :2, ::r..::/r koNmyunike 'communique' and 7 7 -t!-tJ 1) akusesarii not 7 7 ·t: 'Y -tjakuseqsarii 'accessory'. Nevertheless, the rule goes on to list exceptions, presumably based on accepted usage, where mora obstruent /Q/ insertion is acceptable. These include;\ 'Y 5- baqtaa 'batter (for cooking, in baseball): ~/ ·'r 'Y 5 ..... saqtaa 'shutter' and :fJ Y ·r --{ > ~7' kaQtiNgu 'cutting'. Rule 5 seems to have had no effect on subsequent orthographic practice. Donor words with both double (mm) and double (nn) are still often spelt with an inserted mora nasal /N/ (58, 65), while the situation with inserted mora /Q/ is still complex (§3.2.2.2). Rules 6 and 8 deal with palatal glides. Rule 6 recommends writing the palatal glide portion of the katakana digraph reduced. Thus, ';} ..:V 7~ not -::.J..y 7:' jazu 'jaZZ: Rule 8 advises using 7 (a) rather than -1' (ya) after the front vowels /i/ and /e/. Thus, ~ 7 _!piano not 1:::' 'V ./ piyarzo 'piano' and '"7 ~ >' heap iN not ,~-,, 'V ~ ~/ heyapiN 'hairpin'. A greater number of exceptions are listed (e.g.~ --1'V taiya 'tyre: ~=-'V beniya 'veneer') than are illustrations of the rule, and spelling on this point is still unstable today. Rule 7 states that the choonpu - should be used to indicate long VO\\I·els and that ~--t) and _:J>( should be avoided and ;.!··----- and .I.·----- written instead.
·-r
1)
......
Chapter 5. Orthography 175
Exceptions are nevertheless cited for the latter (e.g. <""\-1' :::/ r· peeNto 'paint'), presumably based on accepted usage. Rules 9 to 14 deal with the spelling of innovative pronunciations and all rec-· ommend a conservative approach. The first of these, Rule 9, is difficult to interpret, however. It recommends that 'originaf !- 7 /tu/ and F '7 /du/ be written (to) and r (do). None of the examples cited(-!.£>}- JV"'<' ::-1 zeNtorumaN 'gentle-· marl, / I.-~- Y }- 7 A ]-- bureeNtorasut o 'brain trust: 7 .-{ -t doraibu 'drive: etc.) actually have ltu du tu dul, or anything that may be construed as 'original' !- 7 or F 7, in the donor word. In all cases, the lui is epenthetic (§3.2.2.1). What this rule presumably sought to recommend, then, was that an epenthetic /o/ be employed after donor It! and ldl rather than the epenthetic /u/ often found in earlier gairaigo (136-137). That this rule is based on an infelicitous analysis and is seriously confused is corroborated by the exceptions noted later: 'Y- ~-A cuupiisu 'two-piece (dress): '.Y 1) - curii 'tree: ;:;(•;; -7 zuQku 'canvas, sackcloth' from Du. doek, and ;:;(o-7- zuroosu '(women's) drawers'. Only half of these, cuupiisu and doek, are adaptations of donor }- ".J or r' ".J • Rule 10 is clearer:f and v-row kana should be written with h-and b-row kana, respectively. Thus, t! ?,-;: ~ Y bitamiN not ~ .-( -).-< ~ ~/ vitamiN 'vitamirl, :/7 ~'J }· ll\- b. puraQtohoomu not /-7 'Y I' 7 ;t- b. puraQtofoomu 'platform' (31---39). Noted exceptions, where 'there is an awareness of the donor sound: include 7 .x:. ::~.~.:.:..z 1- feminisuto 'lady's man' and ~;t;:j:.-~, ';17 1) - vokyabu.ra.rii 'vocabulary'. Rule 10 is unusual in having an explanatory f(lotnote which states that at the general meeting of the Second National Language Council there was debate con·· cerning variations in spelling such as 7 .f JJ; };:, .fin.tmu or 7-1 JJ; L fuiru.mu 'film.'. In other words, while not stated in explicitly linguistic terms, whether spellings should reflect a mono- or bimoraic pronunciation. Rule 11 states that ·r ...( (ti) and ~r' (di) should be written 'whenever possible' as ::f- (ci) and ':) {ji), respectively. Thus 7--l;:, ciimu not 7 .f- L tiimu 'teani, -::.'J...,~/''(' jireNma not r-1 [../~/''(' direNma 'dilemma'. Exceptions, where 'there is an awareness of the donor sound', are 7-1- tii 'tea' and 1::: JVr -1 > y· birndiNgu 'building'. See the discussion centred on (17 19) for further detail on the diachronic background to this recommendation. Rule 12 urges, in a similarly conservative vein, that V J: (se) and / .:r:. {je) be written 'whenever possible' as -l! (se) and J.;Z' (ze), respectively. 1hus, ~;J;?-1:! :-t- mirukuseeki not ~ Jj; 7 Y .x. -.:--t· miruk-uSeeki 'milkshake: -1:!7, 1- -1"'- zesulaa not :J .X. A 7--t,····· jesulaa 'gesture' (43, 53). Here, exceptions cited include Y .x. •···· F seedo 'shade' and :J.x:. 'Y 1--.:r.:.Y/Y jeQtoeNjiN 'jet engine'. Rule 12, like Rule 10, contains an explanatory footnote. This makes it clear that the Council did not agree witl1 the Subcommittee's recommendation on this point: '/se/ and /je/ are
r
176
Loanwords in Japanese
sounds that can be pronounced by Japanese people and thus the spellings "/ .:r~ and :J .x. ought to be recommended'. Rule 13 recommends monomoraic 1) 1 { F ·1' ·~' =f· saNdoiQci 'sandwich: not saNdouiQci. A f()otnote reveals that the Council failed to agree with the Subcommittee's recommendation here too: 'depending on the donor sound, there is surely no problem with writing 1) 1 (wi), 7 . I {we) and 7 :t (wo)'. .For further detail on moraicity issues here, see §3.1 and (77). Rule 14 states that the labial glides ~7 T (k\·va), -1 (kwi), /-,.:::. (kwe) and ·) ;t {kwo) be written tJ (ka), 7 ,( (kui) ·)I (kue) and ::1 (ko ), respectively. Thus, 7..-? .J: 7 sukuea not 7..-? x. 7 S'~Akwea 'square: v-f" > ;<.;b "J ~ .2 remoNsukaQ.~u not 1...-'-f: >.A 7 7 ~~, > ::1. remoNsukwaQSu 'lemon squash' (78-80). This recommendation may be seen as inconsistent. While monomoraic 7 1 (k\-\'i) and :7 J.: (k-we) should, like 1) -( {wi) and ~:7 .:r:. (we) in Rule 13, be written bimoraically as ~; ...-( (kui) and ~;.I. {kue), monomoraic ·) 7 {kwa) and ~; :t {k
Chapter 5. Orthography 177
sucuwaadesu not _7, 7 .:1. lJ--r;7.. sutyuwaadesu 'stewardesS, ~ .:1.- :x., juusu not 7 .;::1.. .A dyuusu 'deuce' (84-85). Only one exception is cited: 7'o7.:1. -----·1}'----purodyuusaa '(movie) producer'. (fyu) and i'f .::z. The final rule, Rule 19, is similar in its theme: innovative 7 (vyu) should be written t::: .::1. (hyu) and !:~' ."1. (byu). 'Ihus ~ .:1.- ; ( hyuuz-tA.not 7 ::r. - -~<'fyuuzu 'fuse: -1· > -;...z. ~ .;::1..- iNtabyuu not -1' > ?< i'} .;::1..- iNtavyuu 'interview' (31-39). An addendum to Rule 19 states that the five katakana :rJ: (wi), ..Y. (we), 7 (wo), ':)' (du) and~ (di) should not be used to write gairaigo. A final explanatory note appearing after Rule 19 concerns a point of punctuation. The Council notes that while three different marks have conventionally been and employed to indicate a word break between two gairaigo, the nakaguro of matter the on decision any defer to elected has it =, and two types of hyphen which of the three to recommend. It points out that the use of the nakaguro may be confused ""ith its other major function, roughly equivalent to the use in English of a slash/. Subsequent to these rules, Monbush6 (1955) then goes on to list 17 pages of gairaigo orthographic examples. The report concludes with an appendix containing orthographic examples of foreign names and a list of kana recommended for their transcription. This is broadly the same as those recommended for gen-eral gairaigo orthography; although includes five additional katakana digraphs, one of which is the curious "Cl .::1. (wyu), cited in the Subcommittee's transcription of'Wiirttemberg' as ? .:1. Jl/r :;_;..-·"\;1,7 wyurutmberuku. These five additional digraphs are coded MP in Table 5.2.
Nearly three decades after Monbusho (1955) was published, the National Institute for Japanese Language published an analysis, NINJAL (1984), of the orthographical data found in one of their earlier surveys, NINJAL (1962). This has direct relevance to Monbush6 (1955) in two ways. Firstly; NINJAL (1962) contains data on a range of magazines published in 1956, only two years after Monbusho (1955) was submitted to the National Language Council and four years after the Subcommittee's deliberations began. Secondly, NINJAL (1984) expressly analyses the NINJAL (1962) orthographical data according to seven of the general rules set forward in Monbusho (1955). Both these factors offer a valuable insight into contemporary orthographic practice. MonbusM (1955) Rule 1 stated that gai.raigo should be written in katakana. Although NINJAL (1984) agrees that 'katakana was overwhelmingly to the fore: it does list a significant number of loanwords in ~lNJAL (1962) which had alternative non-katakana spellings. Of these, the following words had a token count
178 Loanwords in Japanese
greater than 20 and showed an alternative non-katakana orthography in 25% or more of these: (269)
ff:, em, !I -~-. p., pp. fi~:, td;f'L:., ·K KW :Jltm H /~-*~·i~,
w ~-~
11-.J
H
(
s~
for mirimeetoru 'millimetre for seNi'imeetoru 'c.entimetre' for peeji 'page' for tabako 'tobacco, cigarette' for kirowaQto 'kilowatt' for meetoru 'metre' for hiopu 'hips' for kurabu 'club' for uesuto or wesuto 'waist' for pari 'Paris' f()r iNti 'inch~ for hiQIO 'hit'
(31/31, 100%) (357/359, 99%) (98/106, 92%) (24/37, 65%) (13/22, 59%) (49/86, 57%) (14/26, 54%) (8/21, 38%) (34/108, 31%) (14/49, 29%) (6/22, 27%) (5/20, 25%)
Monbusho (1955) Rule 3 recommended the mora nasal /N/ be written:/. The central finding ofNINJAL (1984) is that f(n gairaigo whose donor words contained (mb ), 105 L. (mu) is found where {N) is expected in 21% of tokens and 22% of types. This is in comparison with only 5% of tokens and 11% of types for donor {mp ). Examples of the former were ::I P b.1:::' 7 koromubia or ::I P b. 1:: 'Y koromubiya 'Columbia' in 53% (9/17) of cases, against ::I P ;:,tl:: 7 komNbia in 47% (8/17). Examples of the latter were .:.~=---v b. 7" kyamupu 'camp(ing)' in 3/11 (27%) of cases, against 4--1:' >'':i kyaNpu or ;:t:-·Y;:/7'" kiyaNpu in 73% (8/11) of cases. NINJAL (1984) provides an extremely thorough analysis of Monbusho (1955) Rule 7, namely that the chOonpu be used to indicate long vowels. Use of the choonpu varied across the five long vowels, with the following ranking found (percentages show proportion of choonpu usage by type): /aa/ (95%, 810/854), /oo/ (92%, 563/611), /uu/ (86%, 254/294), /iii (84%, 467/558), /ee/ (71 %, 323/455). Examples without choonpu included A 1J-7' sutaa '(movie) star: :> 2 ;t soo 'show: 7'-- _"1. ry 1) ··;1 ..~;/ cuuriQPu 'tulip: 7 :f--1" sukii 'skiing' and :'r:r:.l:. neemu 'name: Of the non-choonpu spellings, notable were "'J {u) for the second mora of/oo/ in 6% (37/611) of types (e.g. -~/;::! ry;t~,_____ r soobooto 'showboat') and-{ {i) for the second mora of /eel in a much larger 26% (119/455) of types (e.g . .:::: :;, ~ 7 J:.-{ 7-._ nyuufeesu 'new member, new employee'). If these two exceptional spellings for fool and /ee/ are ignored, then the use of nonchi>onpu spellings was more common amongst names: 12% (1 07/895) of types
Note that, even as late as 1984, it appears NINJAL is still basing an analysis on donor mthography, not donor phonology. This is surely indicative of how entrenched dictionary traditions are (§3.2). 105.
Chapter 5. Orthography 179
here, as against 5% (92/1721) for other gairaigo. Differences were also evident according to donor language, although in some cases the number of donor words was small. While words borrowed from English showed non--choonpu spellings in 7% (137/2081) of types, this figure rose to as high as 21% (6/29) for Dutch and 17% (34/202) for l:rench, while falling to as low as 2% (2/110) for German. Dictionary traditions (§3.2) were clearly playing a role here. Finally, a non-choonpu spelling was more likely to occur word-finally (15%, 130/892) than word-internally (4%, 69/1724). Although Monbusho (1955) Rule 8 advised using 7 (a) rather than ~ (ya} after the front vowels /i/ and /e/, the statistics in NINJAL (1984) suggest that, in &pite of the instability at the time Subcommittee debate was being conducted, the majority of gairaigo of this type were in fact already being written with 7 (a) after front vowels. Overall, 70% (192/273) of types evinced 7 (a), although this was more common with /ea/ (86%, 42/49) than with /ia/ (67%, 150/224). NINJAis (1984) analysis shows how divorced from the contemporary reality of spelling practice the Monbush6 (1955) Subcommittee's recommendations had been as regards both Rules 10 and 11. Rule 10 stated thatf and v-row should be written \'Vith h·· and b-row kana. The disparity was particularly large with donor I~· Here, the recommended h--row kana were being employed in a mere 6% (13/214) of types. Bimoraic spellings such as :7 7 (fua) were found in only 10% (21/214) of types, with the remaining 84% (180/214) being monomoraic spellings such as 7 7' (fa). With donor lvl, the split was more even, however. The recommended b--row kane~ were found in43% (149/344) of types, with bimoraic spellings such as <'/7 {vua) in only 4% (13/344) and monomorak spellings such as "'/7 (va) in the remaining 53% (182/314) of types. Rule 11 recommended that 7" -1 (ti) and .::r·.-r (di) be written as 7- (Ci) and :J (ii). Both the innovative forms 7" -1 (ti) and 7'.-r {di) were more popular, although the former, with 51% (70/138) of types, less so than the latter, with 73% (60/82) of types. 106 Innovative forms were conspicuously dominant when word final: 88% (28/32) of types for =p.-r (ti) and 81% (17/21) of types for -:''-1 (di). Finally, Monbush6 (1955) Rule 16 stated that 'final-er, -or and -a1·, particularly from English' be spelt long using the choonpu ~. NINJAL (1984) found this to have been the practice of the time. An overwhelming 96% (273/285) of types borrowed from English final -er, -or and -ar were spelt with a choonpu in NINJAL (1962).
I ignore here NINJAI:s (1984) statistics for the adapted forms 7- /tel, 7"-1 /tee/, 7- ..... /tee/, 7! /de/, 7!/( /dee/ and :;-t..... /dee/. See (20) for background.
106.
18o Loanwords in Japanese
5.:P
Orthographical recommendations II: Bunkach6
By the 1980s, many of the recommendations in Monbush6 (1955) were being ignored, and those of Bunkachc3 (1991) being anticipated, in orthographic guidelines published by several major organizations. These included, as noted in NINJAL (1990: 111···112), the Japan Licensing Authority (NSK 1982), the Japan Newspaper Publishers & Editors Association (SYK 1984) and the nation's only public broadcaster, NHK (1987). Bunk:acho (1991) is a document based on debate carried out during the 17th and 18th National Language Council sessions held from 1986-1988 and 1989-1991, respectively. In 1991, it was submitted to the Minister of Education, announced as Cabinet Notification No. 2 and promulgated as Cabinet Directive No. 1. It was the last document submitted by the National Language Council to become either a cabinet notification or directive: the body was disbanded in 2001. Language policy issues are now officially the bailhvick of the National Language Subcommittee (Kokugo Bunkakai), 107 at the time of writing one of four subcommittees within the wide--ranging Culture Advisory Committee (Bunka Shingikai). The preface to Bunkachi::1 (1991) states that its contents are intended to act as an authoritative guide to gairaigo orthography in the spheres of 'laws and ordinances, public documents, newspapers, magazines, broadcasting, etc:: but that it is 'not the document's intention to infringe on individual orthographic practices in scientific, technical and artistic spheres: nor 'in the spelling of proper nouns, including personal names, company names and brand names'. Neither is it the document's intent to 'contradict or negate gairaigo orthographies established in the past'. The document continues with a list of katakana recommended for gairaigo spelling, divided into two charts. The larger Chart 1 is to be used for 'the general writing of gairaigo and foreign names' and consists of the 115 (di)graphs coded B 1 in Table 5.2. The smaller Chart 2 is to be used for 'the writing of the same when necessary to come as close as possible to the original pronunciation or spelling' and is composed of20 (di)graphs coded B2 in Table 5.2. Taking Chart l and Chart 2 together, for a total of 135 (di) graphs, makes for an increase of five (di)graphs over the total of 130 kana (M, M·, MF) recommended in Monbushc3 (1955). Appearing only in BunkacM (1991) are -/ .-( (kwi), 7 .I. (kwe), 7 :t (kwo), .:;- .:1. (tyu), '/'-{ (ci) and r'f .:1. (vyu), while t/ .:1. (wyu) appears only in Monbush6 (1955).
107. Note, however, the input to language policy provided by NINJAL, especially with regard to current issues regarding the proliferation of gairaigo. See Chapter 6.
Chapter 5. Orthography
The main thrust of Bunkach6 (1991) lies in six 'Pundamental Articles' and 24 'Detailed Articles: .Fundamental Article 1 states that the document treats only katakana spellings of gairaigo, while Article 2 warns that no attempt will be made to judge between two already well-established variants of the same donor word (e.g. ~'u~ "":f guroobu. and -:-f·::j'-;1 gurabu. 'glove'). Where a spelling is well-established, notes Fundamental Article 3, it should be adhered to, the use of different variants across different technical fields being perfectly acceptable. Fundamental Articles 4 to 6 provide more detail on the two kana charts described in the previous paragraph. Chart 1 kana should be used 'for the most part' in writinggairaigo whose level of nativization is high, Chart 2 kana to write words whose degree of nativization is 'not high' or which are 'to a certain extent' gaikokugo. Chart 1 kana should be used if Chart 2 kana are unnecessary: e.g. -{ .:C. (ie) for-{.::::. (ye), '7:t {uo) for '7::t (wo), 1'\ {ba) for '1'7 (va). In the event of it being necessary to employ kana not listed in either chart, no rules or advice are proffered and usage is down to the individuaL Ten examples of such usage are cited and these katakana are coded B• in 'Iable 5.2. The 24 Detailed Articles are divided into three groups: (I) Chart 1 'additional' kana; (II) Chart 2 kana; and (III) the mora consonants, use of the choonpu and other matters. Articles I.1 to 1.6 illustrate the usage patterns f()r all Chart 1 kana. I restrict listing examples cited to a small selection of those containing nontraditional (i.e. unshaded in 'Htble 5.2) katakana digraphs: ~ .:r:. ~ t.;~ seekaa '(cocktail) shaker: ~'-·( :J .:r:.'A r daijesuto 'digest, rounduP, "1- .:r:..7-. cesu 'chess: 7:':::.·---- dizu·-· ~/-'(' 'Y'Y .:r:. saNce 'ski jump: )j~ ·} >T .-( J' boraNtia 'volunteer: nii 'Disney: 7 .:r:. 'Y::f/ -:-ljeNsiNgu 'fencing (sport)' and ::1. r dyueQto 'duet'. Also given, in some cases, are exceptions where usage has already been established. For 7" ..-( (ti) and -;" ..-( {di), these include exceptions using both i-column (e.g. ·1- (ci), Y' (ji)) and e-column (e.g. ·r (te}, (de)) katakana (17, 19-20). For thef-row, exceptions employing both h-row (e.g. l'\ (ha)) and bimoraic spellings (e.g. 77 (fua)) are given. Articles ILl to II.lO, which deal with Chart 2 kana, are a simple list of examples. Each of the 10 articles uses wording such as 'the kana 'J' J (gwa) should be used to represent the foreign sound -:lr {gwa)'. The 'foreign sound' is written in katakana and IPA is not employed. Examples cited include -'( .:r:.- ":! yeecu '(William Butler) Yeats: _;;z? .:r:. -·r'~/ suweedeN 'Sweden:: ? :r. ,7., =f- ::1 >''·?~ 7 kwesucoNmaaku 'question mark: J"\5 ';fy-{ paragwai 'Paraguay: "') ;].;._:) .J:. :::.~ 'Y-'( > sorujeniiciN '(Alexander) Solzhenitsyti. l\1--'r }- '!7 1) ·'f> halatrlriyaN '(Aram) Khachaturiari, 7::1.- J "\ tyuuba 'tuba, 7 .:r.. ~ :/ ::;r >' fyuujoN 'fusiori, Jv-l:t -{ .::2 verusaiyu 'Versailles' and ]....-~ ::1. ~ revyuu 'revue, (book) review'. Remarks regarding exceptions note that foreign ~( x (ye) is 'generally written' either as .:r. (e) or --{ .:r. (ie) (82); that the convention of writing, for example,
r
"':r.
r.(
181
182
Loanwords in Japanese
monomoraic l) -1 (wi) rather than bimoraic l) 4 (ui) is especially strong with names (77); that donor 7 7 (kwa) and 7' 7 (gwa) may also be written with a different reduced kana, as ~7 'J and ,Y'r; (78-79); and that foreign ]-- ') (tu) has established spellings with both 'Y (cu) and r (to), likewise F' 7 (du) with both ;;(' (zu) and F (do) (136-137). Of the third and final group of Detailed Articles, JILl and IIL2 treat the mora consonants. Article IIL2 recommends that the mora obstruent /Q/ be written reduced, as ':/-.Article IIL1 recommends the mora nasal /N/ be written >,but that established spellings such as V .b.>-J~:Jr) 1,.. simupojiumu (rather than .Y:;,t:;:J"Z:/ 1:7 1-:, sinpojiumu) 'symposium' should be respected. The noting of this and other excep-tions is further proof that Rule 5 of Monbusho (1955) had tailed to take hold. Article III.3 states that, 'as a rulC, the choonpu should be used for long vowels: e.g. if~-·7 > f' pooraNdo 'Poland'. Ho\\'ever, it notes that there exist established spellings with 0--row kana, especially when these represent /eel and /ooi: e.g. ]/....( 7~' }, reeauto 'layout' and 75::-t#? Jv samdabooru 'salad bow[ loans whose English donor words have final-er, -or or -ar should be written long (e.g. ::I> l:-::' :;. ·-----?- koNpyutaa 'computer'), except where established spellings or variants exist (e.g. :::I :::../1:: .'2. ------~ hmpyuta 'computer'). Articles IliA and III.5 are fundamentally identical to Rules 8 and 17 of Monbusho (1955) and cover the use of 7 (a) after front vowels and the spelling of donor word-final ((i)um). Article IIL6 adopts a more flexible position than that taken by Rule 15 of Monbusho (1955). Whereas the latter advocated an epenthetic /u/ following the lkl in a donor iksl cluster (e.g. -9 7 ';/- takusii not ?<:;f.:/takiSii 'taxi'), Bunkacho (1991) states that an epenthetic /i/ is also acceptable. Current usage is confused (123---126) and, as (124) makes clear, some gairaigo have both variants. Article IIL7 recommends that kana indicating palatal glides, -1', :;;__ and 21 , be written reduced, as should the second kana in the digraphs if 7 (va), ~}-( (vi), ~lJ: (ve), ~f7]-· (vo) and 1~ ':7 (tu). Oddly, this advice is not offered for the latter digraph's voiced counterpart F ') (du). The final Detailed Article, III.8, addresses the issue of how to indicate a word break in a loan phrase. While this was passed over in the fi)rm of a final explanatory note in Monbusho (1955:9), Bunkacho (1991) sidesteps the problem completely, by stating it has no position to take and no advice to offer, other than 'conforming to accepted pmctice: Tims, keesubaikeesu 'case by case' may be written with spaces (7-7--. J\-1' -7-7--.), with nakaguro (--7- ;;z ;\.{ • 7- /Z), or with hyphens (-7- .::Z-l \ 1-7- .::Z), 108 while non -kanji-sphere foreign names
'Ihe commonly found lJ'----- ;:zr\--{-':J'-----7., Vl'here the word break is simply ignored, is not mentioned.
108.
Chapter 5. Orthography 183
(e.g. maruko paro 'Marco Polo') may use a nakaguro ("? JL 1 ::1 ;1o) or a double hyphen(""'"< Jv:::::J =J.:l~O), but, curiously, not a space("';' Jv::J ;J~o). It is worth mentioning one survey whose results are relevant to the Bunkach6 (1991) orthographic recommendations, specifically the use of b-row or v-row kana to write donor lvl. Sakamoto (2002: 55-56) analysed transcription patterns for the V··row in three media: usage found between 1.985 and 2001 in three daily national newspapers (Asahi Shinbun, Yomiuri Shinbun and Mainichi Shinbun); usage patterns derived from the search results of three major internet search engines (Yahoo!, goo and Google) carried out in 2002; and a survey carried out by NI-IK in the same year which asked just under 1,300 respondents which of two transcriptions they usually used. His results are summarized in Table 5.3.
Table 5.3. v-row transcription patterns across three media (adapted from Sakamoto 2002)
beetoobeN: beetooveN 'Beethoven'
baioriN: vaioriN 'violin' ~~-{ 7.1" 1) /: ~7 ··( ;;t 1) /
x:- J- -'"(:/': ·'~- }'_,.X. ~I
Asahi
93:7
94:6
Yomiuri
90:10
91:9
Mainichi Yilhoo!
92:8 49:51
85:15 48:52
goo Coogle
52:48
47:53
52:48
42:58
NHK
75:24
80:19
Transcription patterns fcJr the v-row differ radically by medium. Newspapers are the most conservative, in general following the older Monbusht:l (1955) recom·· mendation to use the b-row. The most politically liberal of the three newspapers, the Asahi Shinbun, is in fact the most transcriptionally conservative, although Sakamoto's figures include orthographic practice found in newspaper editions as far back as 1985 and predating Bunkach6 (1991). Internet search engine results show the most innovative transcription with a rough 50:50 split between v- and b-row usage. The NHK survey falls approximately halfway between these two extremes.
Compared to Monbusht:l (1955), the tone of Bunkach6 (1991) is more forbearing and its content less confused. Although it provides useful guidance on
184 Loanwords in Japanese
how to transcribe gaikokugo, it stops short of offering suggestions for a number of common foreign phones still without an accepted kana transcription. Restricting examples to consonants, these include IJ (3 '3 xl, conventionally transcribed identically to donor lr s z j hi, respectively (Table 3.2). When, therefore, it is 'necessary to come as close as possible to the original pronunciation or spelling' (Bunkachi) 1991: 3--4) a transcription is often difficult or ambiguous when restricted to kana, a situation which has far-reaching consequences for foreign language pedagogy. This is not to say, however, that such transcriptions have not been proposed or attempted. A handakuten on the g-row ( )J" etc.) may be used to indicate a velar nasal, for instance. 109 TI1e match between kana graphs and the sounds they actually represent in modern standard Japanese is close to 100% transparent. This means that a document like Bunkacho (1991) is not only recommending a given spelling, but also the pronunciation on which the spelling is based. When, f()r example, Detailed Article III.3 notes that the word for computer can be spelt as either :J / ·t: ::1. ::>< ,____ koN-pyuutaa or ~-1 >1 ~ ~ ~ -;: koNpyuuta, it is in reality acknowledging that the word has two different pronunciations, one with a final short and one with a final long vowel. Many of the orthographic recommendations in Bunkacho (1991), as we 11 as in Monbusht) (1955), double theref()re as recommendations on gairaigo pronunciation and offer tacit acknowledgement. especially in Bunkach6 (1991), that these often vary. Purely orthographic alternatives among modern Japanese gairaigo alternatives unconnected with variant pronunciations are found in only four areas. TI1e first of these, variation regarding a \Vord break in a loan phrase, is dealt with in Detailed Article III.8. The second, the question of the extent to which the reduced graph in a digraph truly indicates monomoraicity (e.g. 7 ,:r__ /ue/ versus t) .J: /we/), is considered in Fundamental Article 5 (see also §3.1). TI1e third is the use of the choonpu to mark a long vowel, taken up in Article III.3 ofBunkach6 (1991). Examples found in NINJAL (1962), such as 7, 57 sutaa 'star' and :2? 1) ·-;1 7 cuuriQpu 'tuliP, where tautomorphemic /aa/ and /uu/ are spelt without a chiionpu, would, if encountered today, probably be considered nonstandard. There are still, however, a very small number of examples in the modern language where tautomorphemic /iii may be found written without a choonpu. These include ::: .;( ';:7 miira 'mummy' and :t 1) .;( ;IJ yuriika 'eureka'. With !automorphemic /oo/ and /ee/, however, the situation is different. Here, the second mora of /oo/ may be written with l) (u) (270a), though never with ;d-· (o), while the second mora of /ee/ may be represented by either/( (i) (271a) or, more rarely;
e
=r
109.
See Umegaki (1963: 142-143) and Nishimura (1995: 31-32) for other examples.
Chapter 5. Orthography 185
(e) (271b ). Both tautomorphemic /oo/ and /ee/ may also, of course, be written with a choonpu (270b, 271c). .:I.
(270) a.
71 /'jJ-.
b. ~'lJv:I- .A
(271) a.
r ;<..f >
b. ;\]/.X. c. J~l... .-
gurukoosu
'glucose'
-1;!-•:77,~- sausupoo
'southpaw'
domeeN '(web) domain' i.--·.-f/' reepu 'rape' 'ballet' v-Y.:r. regee 'reggae baree 7'-17 deeta 'data' 'volleybalf baree
In a very few cases a different spelling serves to distinguish two homonyms: e.g. bom·iNgu :il;- 1) / ' ~7' ~oring, drilling: but bom·iNgu :il;'71} >~/'ten--pin bowl-ing'. In careful pronunciation, however, type (270a) and (271a) spellings are typically sounded out as /o-u/ and /e-i/ rather than /o-o/ and /e-e/. Vance (1987: 13) comments on an identical phenomenon in Sino-Japanese vocabulary. While I am unaware of any statistical studies offering corroboration, there is a tendency for type (271a) spellings to occur more frequently in names. The homonyms :;<-{ >· meeN for 'Maine' but :J- > meeN for 'main, ---~-{ -;:; peeji for '(Jimmy) Page' but .~~--/ peeji f()r 'page: and quasi-homonyms l-/-{ "/ geecu for '(Bill) Gates' and -'l- }- geeto for 'gate: are all cases in point. The fourth and final area where orthographic variation is apparent is the choice of which script to employ when writinggairaigo. One of the recommendations made by Monbusho (1955) had taken hold so strongly by the late l980s, that mention of it in .BunkacM (1991) was no longer deemed necessary. Bunkachi::l (1991) states nowhere thatgairaigo should be written in katakana, simply noting in Fundamental Article 1 that it will treat only katakana spellings. By 1991 this was the script being used to \\Tite the overwhelming majority of gairaigo. While nearly two decades on from Bunkacho (1991) hiragana is rarely encountered in gaira.igo orthography; a small number of kanji spellings are still fre~.:1uentlyused, as (254-255, p. 168) attest. It is, however, a foreign script, the Roman alphabet, which has made greatest inroads into the katakana hegemony and it is to this that I turn in the final section of this chapter,
5·4 Roman script Although no examples ofgairaigo written in Roman script occur in Asahi Shinbun editorials until the 1940s (Hashimoto 2010: 141), by the 1990s those written in Roman script accounted for 4% of all gairaigo. This had increased to 5% by the 2000s. As schematized in Figure 5.3, Roman script is employed in three major ways
186 Loanwords in Japanese
in modern Japanese: to write acronyms, to write abbreviations, and to write full words. Abbreviations and full words are examined in §5.4.1, acronyms in §5.4.2.
ROMAN SCRIPT
ACRONYMS
ABBREVIATIONS
GAIRAIGO & GAIKOKUGO
/
FULL WORDS
NATIVE & SINO-JAPANESE
l~
,B-0-R-RO_W_ED--,
SEMANTICALLY
'IN-NO_V_A_T-IV-E--,
DIRECTLY (TYPE 1)
REMODELLED ENGLISH (TYPE 2)
(TYPE 3 )
~ NAMES OF CORPORATIONS, ORGANISATIONS, ETC.
,---------,
'KY JAPANESE'
Figure 5.3. The use of Roman script in modern Japanese
5.4.1
Abbreviations and full words
Abbreviations written in Roman script are typically not spoken as letters, but as their expansions: e.g. the Roman abbreviation mm is read mirimeetoru or miri, not *emuemu; TV is read terebi, not *tiibui; kg is read kiroguramu or kiro, not* keejii. Full words written in Roman script are especially prevalent in signage (Someya 2002; Backhaus 2007), newspaper and magazine advertisements (Haarmann 1989), pop lyrics (Ito 2003), brand names, band names and computing (e.g. URLs and e-mail addresses). These words may be from any vocabulary stratum. Frequently, what appear to be gairaigo examples may be better viewed as gaikokugo: e.g. Q, k, m, n, o) in Table 1.2 on page 12. Various romanization systems exist, the principal of which are the hebonshiki """~ /' :t. 'Hepburn style' and the kunreishiki "tJll1'- :t. 'cabinet-ordered style' transcriptions (Unger 1996:50-56; Seeley 2000:138-142, 205-206). 'Recent borrowings pose some serious challenges' (Vance 2008b: 87) for kunreishiki, and hebonshiki is best suited to writing gairaigo in a manner reflecting the contemporary phonemic system in Figure 3.2. Table 5.4 shows contemporary orthographic practice for the 20 most frequent gairaigo extracted from NINJAL (2006a), a survey of magazines published in 1994
Chapter 5. Orthography 187 Table 5.4. Orthographic practice for the 20 most frequent gairaigo in magazines (extracted from NINJAL 2006a) Proportion
<~uraigo
katakana
Romansnipt abbreviation
mirimeetoru seNCimeetor u
meetoru
guramu
-~) .~
··---- }- JlJ
<1%
em
1:: ~/+ i-- ,"- Jl,·
97%
3%
m
J<-}·Jv 7%
93%
type
.;::: ..,· 7' 96%
4%
set
1::';1
99%
1%
tel
~r L-"*~.IJ-r 1..··7 ;j· :/
97%
q-_ 0
km
3% _:;.\ ·--- }- Jl-·
93%
7%
g
77 L.
saizu eN} iN
:3.
>99%
seQ to
kiromeetoru
<1%
>99%
mm, m/m
taipu
tere
full word.
4%
96%
size
"7-1'7;'
2%
1%
98%
eng
engine
.T... ~/~Y::/
moderu
6%
9,!%
model
·t::·r)t/
hoteru
4%
96%
hotel
;$-J-/L-·
karaa dezaiN
sisutemu supoocu peeji suupaa
4%
96%
c.olor
t;.::J.---
4%
96%
design
7"-P;l'-{ ~/
5%
95%
~ystem
"./7-'.''j-];,
ll%
89%
sports
A;.l~-----'Y
2%
57%
41%
page
«·-----;:/
""!F.f
14%
86%
super
;:z-;~-
188 Loanwords in Japanese
(§1.3). The gairaigo are listed in frequency order and each may appear in any or all of three scripts: Roman, katakana or kanji. Those written in .Roman script may appear as an abbreviation or a full word. None of the 20 most frequent gairaigo appear written in hiragana. Altogether, a total of seven gairaigo in the Table (tere and faQkusu, plus five measurements) have Roman abbreviation rates of over 90%. The only gairaigo in the Table to appear with any great frequency as a full word in Roman script is za "' ji. Peeji 'page' is the only loanword in Table 5.4 to record may still occasionally be found tokens written in kanji, although ft J1 and for mirimeetoru, seNCimeetoru and meetoru, respectively (269, p. 178). Although NINTAL (2006a) also analyses the proportion of tokens appearing in magazine advertisements, a medium where Roman alphabet gairaigo are more likely to be found, there appears to be no correlation between this proportion and that of full \·vords written in Roman script.
*·
5.4.2
Acronyms
As illustrated in Figure 5.3, Roman acronyms may be divided into two groups based on their expanded forms: gairaigo and gaikokugo; and native and Sino-Japanese. The gairaigo andgaikokugo Roman acronym group may be further subdivided into three types: borrowed acronyms (type l), semantically remodelled (SR) English acronyms (type 2), and innovative acronyms (type 3). A selection of these three types is shown in Table 5.5. Overwhelmingly from English, borrowed acronyms may be spoken in two different ways: as letters (type la) or as a 'Nord (type lb). Occasionally, Japanese usage differs from the donor. UFO, for example, is typically spoken as a three·· letter string in English but always as a word, yuufoo, in Japanese. When spoken as letters (type la), it is adaptations of the English letter names which are typically found. One exception is the alternative peehaa for pH, adapted from the German. Some type l acronyms are considerably more frequent in Japanese than they are in their donor language. DV, NG,AVand NEETare cases in point, being relatively infrequent in English. ·with the demise of the video tape, VTR has lost currency in English- in Japanese it is now typically used in the sense of'video cli{i. especially in TV programmes. SR English acronyms (Type 2) do not exist in English and have been created in Japan from English word strings. Many are corporate names and may contain a letter with a Japanese expansion, often N for niopoN 'Japan'. Like the SR com·· pounds in (246···251) on pp. 156···157, SR English acronyms can be a cause of language learning problems. Innovative acronyms (Type 3) are composed of a mixture of initials and other word segments. In most cases the innovation took place in Japan: e.g. the
Chapter 5. Orthography 189
unattested acronym *lvlECSST would have been read as a word mekusuto, now written in .Roman capitals as the faux--acronym MEXT. In a few cases, the innovation has been borrowed: e.g. XL. Table 5.5. Selection of Roman acronyms in Japanese Acronym
Pronunciation
Source
la 1a la 1a 1a la la la 1a 1a Ia 1a 1a 1a la 1a la la 1a
4WD AP ATM AV CD DV EU FM IT
yoNda buryuudii eepii eetiiemu eebu! Siidil diibui !iyuu efuemu aitii er1ujii enujiioo oobii piieeCi, peelzaa pi!aaru
WTO
kyuuclNdoee esuooesu buitiiaaru daburyuutiioo
4WD (four-wheel drive) AP (Associated Press) ATM (automated teller machine) AV (adult video) CD (compact disc) DV (domestic violence) EU (European Union) FM (frequency modulation) IT (information technology) NG (no good) NGO (non-govemmental organization) OB (old boy) pH (Ger. Potenz H 2) PR (public relations) PTA (parent~-teachers a.~soclation) Q&A (questions and answers) SOS (save our souls) VTR (video tape recorder) WTO (World Trade Organization)
1b
AIDS
eezu
lb
NEET
niito
1b
TOEIC
tooiQku
lb 1b
UFO UNESCO
yuufoo yunesuko
2 2
lF 1LDK
itief.t, iQkai waNerudiikee
2 2
ANA B1
eeenuee, ana biiili, bliwaN, biiiQkai
Type
NG NGO OB pH PR PTA Q&A
sos VTR
piitiiee
AIDS (acquired immune-deficiency syndrome) NEET (not in employment, education or training) TOEIC (Test of English for International Communication) UFO (unidentified flying object) lJNESCO (United Nations Education Scientific & Cultural Organization) 1 floor ("" ground floor/first floor) (apartment having) 1 bedroom+ living room, dining room & kitchen All Nippon Airways basement 1 (=first lower ground floor)
(Continued)
190
Loanwords in Japanese
Table 5.5. Selection of Roman acronyms in Japanese (Continued) Acronym
Pronunciation
Sow·ce
GM
.siiemu
2
IC
aisii
2 2
JR MOS(burger) NEC NTT OL PK TBS
ieeaaru mosu
commercial message(= TV advertisement) (motorway, expressway) interchange Japan Railway (sic.) mountain ocean sun Nippon Eledric Company Nippon Telegraph & Telephone office lady penalty kick Tokyo Broadcasting System
T}'PC 2
2 2
2 2 '"I
"'
enuiisii enutiitii
ooeru piikee tiibiiesu
3
MEXT
mekusulv
3 3
TO PIX XL
eQkusueru
topiQkusu
Ministry of Education, Culture, Sports, Science and Technology Tokyo Stock Price Index extra large
Acronyms across all these three types may contain numerals or other symbols. The numeral portions are typically read in Japanese (e.g. yoNdaburyuudii, not "foadaburyuudii, for 4WD; iCie.fu, not "WaNefu, for JF) or sometimes in English also (e.g. biiiCi, but occasionally biiwaN, f()r BI). Examples where the numeral portion is read only in English are few, but include waNerudiikee, not *icierudiikee, for lLDK. A Roman acronym may possess an alternative katakana spelling: e.g. ~~ 7~ Jv f(Jr PR. Others may appear in hybrid compounds with mixed orthography: e.g. PK~ piikeeseN 'penalty shot-out: AB~ eebiigata 'AB blood group'. Statistics on the most frequently found gairaigo Roman acronyms may be extracted from Nakayama et al. (2007: 377), an analysis of all Tokyo editions of the Mainichi Shinbun published between 1994 and 2003. The ten most frequently recorded were, in descending order: JR. NTT, EU, AP, TBS, NBC, NGO, WTO, CD and IT. The frequency rates for these acronyms are revealing in that they underline just how important the role of the Roman alphabet has become in writing gairaigo. If the most frequent loanword in Nakayama et al. is normalized to 100, then ]R has a frequencyof31 and ITa frequencyof9.]Rwould rank as the 17th most frequent and IT as the 115th most frequent gairaigo in their database. 110 I close with a brief review of the native and Sino--Japanese Roman acronym group. As illustrated in Figure 5.3, such acronyms come in two main types. Firstly,
no.
When foreign names are ignored. For greater detail on this survey see Table 2.11.
Chapter 5. Orthography
the names of corporations, groups or organisations. Here, a Roman acronym typically serves as a display of stylishness or international savoirfaire: e.g. NI:lK for niQpoN hoosoo kyookai 'Japan Broadcasting Corporation'. The second type is what has come to be known as 'KY Japanese' (kewaishiki nihongo KY:i\. 8 :1jr.[it}) after the acronym KY f()r kuuki yomenai 'out of touch with realitY, lit. 'can't read the air'. This acronym came to media prominence in 2007 to describe the cabinet of the then prime minister Abe Shinzo 5i:·1t-~ =.. KY Japanese is used most frequently by the youngest age cohorts, especially in texting (SMS messaging) and, as the origin of the term suggests, is often pejorative. Examples include IW for imi wakaranai 'don't know what you're talking about', and DSK for debu sukebe kusai 'fat, crude and smelly'. KY Japanese may also be fixmed in innovative fashion: e.g. 3M for majide moo muri 'totally impossible'. Some examples may contain initials corresponding to a gairaigo: e.g. FK for JaNde koi '(overly) heavy make- uP, where JaNde is a back-clipping (§3.4) of jaNdedoN 'foundation'. For further examples of KY Japanese see Kitahara (2008). Although predating the coining of the term 'KY Japanese. perhaps the most commonly encountered non-gairaigo Roman acronym is H, 'sex, dirty, lewd: pronounced eQci, from Sino-Japanese heNtai 1t ~ 'pervert'.
191
CHAPTER
6
Attitudes to loanwords
1he sense that the Japanese language is succumbing, or has already succumbed, to a 'deluge' (the Japanese term most frequently employed is haNraN) of foreign loans appears all-pervasive. Criticism has long been especially fierce with regard to the amount ofgairaigo employed by the bureaucracy (see, for example, Mogami 1984). Research carried out by Yamada & Nanba (1999) has shown that, while the number of letters and articles critical of gairaigo in the pages of the Nih on Keizai Shinbun and Asahi Shinbun were relatively few in the 1950s, they rose dramatically in the 1960s and peaked in the 1980s. Honna (1995: 45-46) comments on the letter pages of the national press being filled with readers' complaints about the 'excessive use of undecipherable, unnecessary, undesirable and misleading words' borrowed from English instead of 'beautiful, authentic' Japanese. But is this loanword deluge real? Sekine (2003:31--33), for one, maintains that the public perception of a deluge is an illusion created by the comparatively short lifespan of the average gairaigo and its subsequent replacement by another. As noted in §1.3, when measured by type the proportion ofloanwords is relatively high and continues to rise inexorably Measured by token, however, the proportion is considerably lower, since many gairaigo are of low frequency (Table 1.4), restricted to a specific register, or confined to specialist usage. There have been many studies focusing on levels of recognition and comprehension of individual loanwords, as well as on their frequency of use (Ishiwata 1965; Ishino 1974, 1977:213-218, 1983:186-231; Ishino et al. 1988; Onishi & Kajiki 1995a, 1995b; Loveday 1996: 162-188; BBK 1997,1998,2001, 2002; NINJAL 2004a, 2005b; Mogi 2007). A diachronic comparison is problematic, hm\'ever, the same gairaigo rarely examined more than once and respondents t)'lJically coming from different age cohorts, socio-economic groups, educational backgrounds and Japanese regions. Although I intend to focus in this section on attitudes towards gairaigo, both of the general public, as well as of the government and the bureaucracy, it is nevertheless revealing to examine one recent example of a study on public recognition, comprehension and use. Since 1995, the National Language Division (Kokugoka), part of the Agency for Cultural Affairs (Bunkacho) and charged with matters related to Japanese, has been carrying out annual opinion polls on language use and public attitudes to
194 Loanwords in Japanese
language. Among the many questions posed in each poll, a small number have concerned loanwords. A recent poll, BBK (2008), offered respondents a list of 60 loamw>rds and a series of questions from which each loanword's levels of recogni·· tion, comprehension and frequency of use were measured. In a poll conducted five years previously, BBK (2003), the same questions had been asked of 120 loanwords whose number included these 60. Table 6.1 shows the top five loanwords in BBK (2008) for each of three measurements: level of recognition, level of comprehen · sion and frequency of use. Figures in parentheses show the percentage point difference from the earlier BBK (2003). All loanwords in Table 6.1 are from English unless otherwise stated. Table 6.1. Top 5 most recognised, understood and used loanwords in BBK (2008) Level of recognition
2
sutoresu 'stress' ro•1958] iNtaaneQlo 'internet' [ 1990s]
3
risaikuru 'recycle' [0 1991]
4
sutaqfi< 'staff' [0 1939]
99% (+1) 98% (+1) 97% (+0) 97% (+3)
rekuriee.SoN 'recreatim{ [0 1941 J
97% (+2)
Level of comprehension 1
sutoresu 'stress' [0 +1958]
98% (+6)
2
risaikuru 'recycle' [0 1991]
96% (+5) 96% (+12) 95% (+7) 95% (+4)
3
sutaqfu 'staff' [0 1939]
4
rekurieeSON 'recreation
5
boraNtia 'volunteer' [0 1958]
[0 1941] Frequency of use
2
risaikuru 'recycle' [0 1991]
93% (+2) 90% (+3) 87% (+6)
sutoresu 'stress' ro•1958]
ro 1941]
3
rekurieeSON 'recreation
4
bortlNtia 'volunteer' [0 1958]
85% (·l)
5
teema 'theme' (+l910]
80% (-0)
Only seven different loanwords appear in the three top-five lists, indicating a strong connection between recognition, comprehension and frequency of use. Sutoresu, risaikuru and rekurieesoN appear in all three lists. Meanwhile, although not shown in Table 6.1, the bottom··five 'least recognized: 'least comprehended' and 'least used' lists all contain the same five loanwords, albeit in differing orders: tasukufoosu 'task force: mesena 'corporate support of the arts' ( < h. mecenat), koNsoosiamu 'consortium', eNpawaameNto 'empowerment' and eNfoosumeNto 'enf<-ncement'. In contrast to the loamwHds in the top ··five lists, these five loans have
Chapter 6. Attitudes to loanwords 195
all been borrowed relatively recently. None have attestations cited in NKD and none are listed in Arakawa (1977). Although levels of recognition and frequencies of use for both the top and bottom five loan
An early survey conducted by the Japanese government on public attitudes to
gairaigo was carried out as part ofNSKH (1977) on just under 8,200 respondents. An ongoing series of opinion polls (BBK 1995, 1997, 2000, 2003, 2008) carried out by the aforementioned Agency for Cultural Affairs have approached loanwords in more depth. A 11 the BBK polls were administered nationwide and elicited between 1,900 and 2,300 valid responses. In wide·.ranging surveys of approximately 3,000 people, NINJAL (2004a, 2005b) also posed a series ofloamvord-related questions. Finally, in 2002, the public broadcaster NHK commissioned a survey of just under 1,300 respondents whose results are available in Sakamoto (2002). While space precludes an examination of all these surveys in detail, there are cases where a broadly similar question has been posed in two or more of them, so allowing a diachronic comparison to be made. Such statistical comparison nevertheless requires caution, since in all these surveys the 'Nords gairaigo, gaikokugo and katakanago are frequently used interchangeably and in a non··specialist sense (§1.2.4). NSKH (1977), NHK (Sakamoto 2002), NINJAL (2004a) and BBK (2008) all posed the similar question: do you feel there are too many katakanago/gairaigoigaikokugo in the words you read and hear every day? A clear majority in all four surveys felt there were 'often' or 'sometimes' 'many' or 'somewhat too many' loanwords. This figure is highest for the NSKH ( 1977) survey and lower thereafter (chronologically, 97%, 78%, 86% and 81%, respectively). 'Ihe two most recent, NINJAL (2004a) and BBK (2008), both provide data by age cohort. ]Jere, it was not the elderly, but rather the middle··aged cohorts who felt most strongly there were too many loanwords. NINJACs (2004a) analysis by employment sector is also \·vorthy of note. Here, 96% of those \·\'orking in management and administration thought there \-\'ere 'often' or 'sometimes' too many loanwords, as did 91% of office workers. This was the case for only 68% of those working in agriculture, forestry and fisheries.
196
Loanwords in Japanese
BBK (2000), BBK (2003) and BBK (2008) all carried the approximate question: what is your opinion of using katakanago/gairaigo/gaikokugo in everyday speech? A comparison of the three polls is shown in Figure 6.1, where it can be seen that indifference rules the day. That said, a creeping negativity is also evident, with those harbouring an undesirable opinion having increased from 36% in BBK (2000) to 40% by the most recent BBK (2008).
• BBK2000
40 +-----------------~~~
• BBK2003
30 +--------------i
• BBK2008
20 +------------10 0 Desirable
Undesirable
No opinion
Don't know
Figure 6.1. Opinions on using loanwords in speech and writing (BBK 2000, 2003, 2008)
While statistics broken down by age cohort are not given in BBK (2003), both the BBK (2000) and BBK (2008) surveys show a clear correlation between age and outlook: the older the age cohort the higher the proportion of'undesirable' and the lower the proportion of 'desirable' responses. Respondents with no opinion were considerably more common amongst younger cohorts.
80 ,----------------------------------------------70
60
.~-------------------------------------- •BBK2000
50
40 30 20
--------------------------------
• BBK2003
'-,,----------------------------- • BBK2008
10 0 Some things Loanwords are easy to can only be expressed using understand a loanword
Japanese has Loanwords always been enrich Japanese borrowing language and words culture
Loanwords have style
Figure 6.2. Reasons for a desirable outlook towards loanwords (BBK 2000, 2003, 2008)
Chapter 6. Attitudes to loanwords 197
Those who expressed a positive 'desirable' opinion in each of these three surveys were then asked to choose one or more reasons why. These are shown in Figure 6.2. The idea that certain objects or notions can only be described or expressed using a loanword is by far the biggest factor involved in promoting a positive outlook towards loanwords. A belief in their ease of comprehensibility is also frequently cited, but this witnessed a sharp drop in the most recent poll. Meanwhile, the notion that loanwords are stylish has been steadily decreasing at the expense of an awareness that the Japanese language has always been borrowing. The reasons cited by those who expressed a negative 'undesirable' opinion are shown in Figure 6.3. Here, the belief that the 'intrinsic goodness' of Japanese is being lost and that loanwords are difficult to understand are the two largest factors underlying a negative stance. The first of these two factors, along with another notion, that loanwords corrupt both language and culture, have both been posting steady gains.
70 ,------------------------------------------------60 + - - - - - - - - -----l
so
;- - ---------------------------- • BBK 2000
40
• BBK2003
30
• BBK2008
20
10 0 The intrinsic goodness of Japanese is being lost
Loanwords are hard to understand
Japanese language and culture become corrupted
Using loanwords just seems like posing
I hate loanwords
Figure 6.3. Reasons for an undesirable outlook towards loanwords (BBK 2000, 2003, 2008)
Without asking whether loanwords were 'desirable' or 'undesirable', NHK (Sakamoto 2002) asked respondents: what do you think when gaikokugo or gairaigo are used? Multiple responses were acceptable. The most commonly heard opinion (50%) was that 'the meaning is hard to understand: followed by the similar 'because they are written in kana not kanji the meaning is difficult to decipher' (34%). Other responses included 'suitable nuances not possible in Japanese can be expressed' (29%), 'Japanese linguistic tradition is being destroyed' (25%), 'the speaker sounds pretentious' (12%) and 'the speaker sounds well educated' (10%). BBK (1995) contained the following question: it has been remarked upon that recently the number of times one comes across gairaigo and gaikokugo in everyday
198 Loanwords in Japanese
life has been increasing. What is your opinion on a further increase? In spite of the fact that the question is leading, a majority of respondents answered that either a slight increase (45%) or an increase of any proportion (13%) would be acceptable. Attitudes were more positive amongst men (62% chose one of these two responses) than amongst women (55%), while negative opinions increased with age (only 34% of women over 60 chose a positive response, while 84% of women aged 16---19 did so). Almost a decade later, NINJAL (2004a) posed the more neutral question: what do you think about a further increase in the number ofgairaigo? Attitudes had hardened in the intervening years. Only 37% felt any kind of increase would be acceptable, a fall of 21 percentage points. Attitudes were once again more positive among men, though only slightly, and once again negative opinions increased with age (only 22% of men over 60 chose a positive response, while 63% of men aged 15-19 did so).It was students \·vho evinced the most positive attitude (63%), \·\'hile on the other end of the scale only 17% of the self-employed and 21% of those working in the agriculture and fisheries sector expressed a positive opinion. When NINJAL (2004a) asked -what do you think should be done about gairaigo in the future? respondents were more sanguine. A full66% felt matters should be left to run their own course, with only 22% stating their numbers should be reduced. Here, however, the difference between age cohorts was stark. Opinion was sharply divided at around the age 40 mark, with 35% of men aged 60 or over wanting to see a reduction, as opposed to only 8% of women aged 15-19. Once again, it was students (71 %) who were most comfortable with letting nature take its course, while those working in the agriculture, forestry and fisheries sector (55%) were the most minded towards intervention. Both BBK (1997) and NHK (Sakamoto 2002) asked the question: do you ever have trouble understanding the meaning of katakanago appearing in ne-wspapers or on TV? Answers were 55% in the affirmative ('often' + 'sometimes') for BBK (1997), but this had risen to 80% by Sakamoto (2002). Women admitted in BBK (1997) to having more trouble than men \·vith loanword meanings (60% 'often' or 'sometimes: as against 49% for men). Not surprisingly, admissions of difficulty increased with age, with sharp rises occurring after the age of 50. A full 67% of women over 60 in BBK (1997) had trouble with comprehension, while this was the case for only 33% of men aged 16-19. NINJAL (2004a) asked: what do you think of so-called 'made in Japan loanwords (waseigairaigo)' such as maihoomu 'owned (not rented) home', pasokoN 'personal compute1'' and koNbini 'convenience sto1'e'? Here, 62% of respondents were positive, with women (64%) more so than men (60%). As is to be expected, attitudes grew negative with age, though these were not as negative as those towards loanwords in general: 50% of men aged over 60 were positive, as against 76% of females aged 15-19. In this question, NINJAL (2004a) made no attempt
Chapter 6. Attitudes to loanwords 199
to distinguish semantically remodelled compounds (246-251) such as maihoomu from truncations (Table 3.9) such as pasokoN and koNbini. NHK (Sakamoto 2002) was more careful to do so. Approximately 55% of respondents said that using semantically remodelled compounds was 'to some extent unavoidable: while 17% felt they should be avoided and 6% thought using them was 'not good: Attitudes to truncated compounds were slightly more favourable, with 58% favouring their use, and 26% against. In NINJAL (2005b) respondents were asked whether Roman alphabet acronyms (§5.4.2) were 'easy to understand: 'convenienf and 'easy to remember'. The proportion of those who agreed was, respectively; 57%, 68% and 66%, with agreement found less among older cohorts. Nearly 41% of over-70s, f(H example, thought Roman letter acronyms were difficult to understand. A slim majority of respondents (51%) felt that the Roman alphabet ought to be officially recognized as a Japanese script, although here the proportion differed radically by age. While over 70% of those under 25 thought this should be the case, this fell to only 39% of 65··-69 year olds and 26% of over-70s. NHK (Sakamoto 2002) asked respondents how they felt about using Roman script to write full words: e.g. soohiN o geQto sum 'get a prize' as(~ Jfr 0.:gef9 0 ). Only 16% felt this was 'good' and a further 46% said they 'didn't mind as long as the words are simple'. A.lthough just under a third overall thought it was 'bad: opinion varied markedly by age: only 15% of 20---29 year-olds, as opposed to 46% of over-60s. Finally, NINJAL (2004a) asked in which semantic domain loanwords posed the greatest problems for comprehension. Here, 59% chose 'computers and related technology: 59% 'government and economics' and 49% 'health and welfare'. Only 3% of respondents stated there was no domain in which they had comprehension problems. 'Ihe same three domains were also the most cited (41 %, 56% and 56%, respectively) when respondents were asked in which areas they would like to see non-loanword alternatives put forward. It is significant that these three domains are perhaps those most important for surviving in and contributing economically to modern society. Being unable to function in any of the other domains offered to respondents, 'fashion: 'music: 'sports' and 'cuisine: poses fewer problems and these received significantly lower scores in both questions. Although the above survey extracts represent only a small portion of the mass of questionnairing conducted in the past half century, two trends are evident. The first is that comprehension of loanwords, especially the more recently borro'~Ned, becomes weaker with advancing age. Such an observation is nothing new, nor is it at all surprising. What is particularly problematic in the Japanese context is that the language of medical care and \·\'elfare for the elderly is replete \·\'ith loanwords (e.g. rihabiri 'rehabilitation: deesaabisu 'ambulatory care, hoomuherupaa 'home helper') due to recent \\'estern-oriented reforms in the Japanese welfare and healthcare
200
Loanwords in Japanese
systems. The second trend is what may be termed a 'love-hate relationship' with gairaigo. Although some view loanwords in a linguistically imperialistic or colo·· nialist light as a threat to Japanese culture and tradition, many others view them as indispensible for creating a more advanced, democratic society. Some view the ever-increasing proportion of loanwords fimnd in daily newspapers and school textbooks as a hindrance to comprehension and learning. Others are acutely aware that the large-scale absorption of Western ideas, technology and loanwords which came with the late 19th century opening up ofJapan played a major role in saving the country from the colonial fate of most other Asian nations.
Within Japan, Sino-Japanese long enjoyed prestige statu<> as the written language of the educated elites. The use of Sino-Japanese in government created an officialese which was so far fi·om the spoken language of the general population that it effectively restricted information access and preserved the status of its users. With the post-war language reforms, Sino-Japanese was eliminated and democratisation of the language enhanced, while the relative prestige of official language was reduced... Since elite groups produce distinctive terminologies, these reforms created the preconditions for the development of a new form of officialese. The current trend for the incorporation of novel gairaigo into the language of the bureaucracy functions to create a new high-prestige dialect which is the preserve of the elite. Tomoda (1999: 250)
On the other hand, [i]f we accept that the motives of ofiicials for using new gairaigo are populist, we are led to the conclusion that the lall),>Uage of officialdom is converging with that of advertising and the mass media at the expense of satisfying the main reason for reform of government language use: the promotion of plain speech. Tomoda (opus cit.: 248)
In a case of one arm of the government investigating another, a question posed in an Agency for Cultural Affairs opinion poll (BBK 1.999) surveyed the degree of intelligibility and level of familiarity of eight different loanwords chosen as representative of those 'appearing in government and municipal office literature'. Responses were received from 2,200 people. Sukiimu 'scheme, plan', akauNtabiritii 'accountability' and koNseNsasu 'consensus' all recorded intelligibility and familiar·· ity rates of less than 10%, while niizu 'needs, demandS, bijoN 'outlook, vision' and risuku 'risk' recorded rates of between 20-40%. Only ibeNto 'event' and meriQto 'advantage, merit' were judged both more intelligible and more familiar than their closest non--loan equivalents moyoosi and riteN. In spite of the fact that 81% (48/59) of prefectures and designated cities had drawn up a manual for improving language use in municipal offices by 1997
Chapter 6. Attitudes to loanwords
(Jinnouchi 2000), a tacit acknowledgement that public perceptions were still negative underlies a more general question on use of loanwords in government oilice literature. This 'Nas posed in an Agency f(n Cultural Affairs opinion poll car·· ried out three years later (BBK 2000): what do you think should be done with respect to gairaigoigaikokugo in order to make information and leaflets at government and municipal offices more intelligible for the reader? Th.e most common response (43%) from the choices oflered was that, while the use ofloanwords was unavoid·· able, their meaning should be explained. A further 39% believed limits should be imposed, 10% stated loanwords should be avoided \·\'herever possible, and a mere 5% of respondents thought they should be actively promoted. Earlier, in 1989, Minister of lJ.ealth and Welfare, Koizumi Jun'ichir() +JR.~- ~15, had set up a committee to look into the overuse of loanwords in government offices and, on being reappointed to the same ministership in 1997, immediately resurrected a defunct Language Normalization Committee (yogo tekiseika iinkai). Koizumi assumed the prime ministership in 2001 and in an alter·· cation with Katayama Tr>ranosuke Jt d-t--.t1L then M.inister oflnternal AlTairs and Communications, during a meeting of the Economy and Hnance Advisory Council in May 2002, stated 'you've got to use words that can be understood rewrite this so that ordinary people can make sense of it' (Tanaka 2005: 8-·-9). Th.e words Katayama had used in his report on information technology had included autosoosiNgu 'outsourcing: baQkuofisu 11ack office' and iNkyubeetaa '(business) incubator'. A similar altercation occurred just over a fortnight later at the same Council meeting, this time with the Minister of Agriculture, Forestry and Fisher·· ies, Take be Tsutomu ~·if~ th, over his use of toreesabiritiisisutemu 'traceability system'. In both instances the ministers concerned had pointed out that replacements or 'Japanese translations' were 'not necessarily possible' or 'very difficult'. Just three months later, in August 2002, the National Institute f(n Japanese Language (NINJAL) set up a Loanword Committee (gaimigo iinkai) with responsibility for producing loanword replacements. Between 2002 and 2004, the Loanword Committee carried out a series of four surveys in order to ascertain the levels of awareness, comprehension and use of 398 selected gairaigo (NINJAL 2007: 265-289; see Tanaka 2007 for a more detailed analysis of the data). These were grouped into four 'star' categories based on level of comprehension: '1-star' loans were those understood by less than 25% of respondents, '2-·star' loans those comprehended by 26···50% of respondents, and so on. The most widely understood '4-·star' loans were removed and the remaining 176 loanwords became the focus of the Committee's deliberations. Replacements for these loanwords were based on six factors (NINJAL 2006c: 18-23): general level of comprehension; level of comprehension among members of the oldest age cohort (aged 60 and above); any diiierence in meaning between a loanword and the word
w
201
202
Loanwords in Japanese
from which it was borrowed; semantic range; level of use in specialist domains; and projected future importance. Recommended replacements were published in a series of fi-mr reports (NINJ AL 2003a, 2003b, 2004c, 2006b ), as a paperback compendium for public consumption (NINJAL 2006c) and as an amalgamated report (NINJAL 2007: 11-261). Table 6.21ists the 20 least comprehended loanwords considered by the Loanword Committee, along with their main recommended replacements. These are all '1-star' loans and thus, by my definition in (4) on p. 10, arguably gaikokugo. None appear in Arakawa (1977) and only two in NKD, both without attestations: U) rdipieNto and (q) ritera5ii. All the loanwords in Table 6.2 are from English sources, shown in the third column. In most cases there has been little change in meaning, although in a few considerable semantic narrowing has occurred: (c) iNkyubedoN refers to business incubation programmes, (f) ofusaitoseNtaa refers specifically to an administrative unit for dealing with public liaison in the event of an accident at a nuclear reactor, (j) rdipieNto to the recipient of a donor organ, and (r) saabeeraNsu typically refers to continuous monitoring by means of surveys and questionnaires. The fourth and fifth columns show the degree of comprehension, both overall and restricted to the oldest over-60 age cohort. In all cases, average comprehension rates among the elderly were lower than those of the general population. In most instances, the Loanword Committee recom-mends only one replacement, shown in the rightmost column, although two are put forward for both (n) eNpawaameNto and (g) ritera5ii, corresponding to different English polysemes. In all cases, the replacements are drawn wholly or partly from the Sino-Japanese vocabulary layer. Those hybrids (§1.2.5) drawn only partly from the Sino-Japanese layer are typically mixed with native Japanese (e.g. riyoosiyasusa for (g) akusdibiritii). 'I\-vo suggested replacements still contain loanwords: (f) geNsiryokuboosaiseNtaa and (t) haisucuzero. Length may be one reason for a negative attitude towards loanwords. Table 6.3 shows the number of characters and the number of moras in both the '1-star' loans and their recommended replacements listed in Table 6.2. In only one case does the recommended replacement consist of more characters than the loan: yomikaki-nooryoku ~.fi- :f -~ ~ir, :)J f()r (q) ritera.5ii l) j· 7 :;, -. Overall, loanword replace-ments offer significant economies tor the printed word. The average replacement is composed of 4.3 characters, as opposed to 7.7 characters for the average loan. When considered from the angle of the spoken word, in only 27% (6/22) of cases does the recommended replacement contain more moras than the loan itself In most cases the replacement is only slightly shorter, however. The average length of the 20 loans in Table 6.3 is 7.7 moras, while that of their 22 recommended replacements is 7.1 moras.
Chapter 6. Attitudes to loanwords
Table 6.2. Recommended replacements for the 20 least comprehended loanwords considered by the Loanword Committee (adapted from NINJAL 2006c, 2007)
'1-star' loan
English source
a roodopuraish'
Degree of {:omprehension (%)
Au.
OVER-60S
road pricing b paburiqkuiNbombumeNto public involvement c iNkyubeesoN incubation d eNfoosumeNto enforcement e koNsoosiarrm consortium
3.0 3.3
1.1 1.-1
3.3 3.4 -1.1
3.3 1.1
f
off--site centre
4.2
3.9
g akusesibiritii
accessibility
4.4
2.1
h akauNtabiritii ajeNda j resipieNto k tastikufoosu
accountability agenda recipient task force
4.4 4.7 4.9 4.9
3.4 3.2 2.7 3.7
kauNtaapaato m koNpuraiaNsu n e!lipawaame!llio
counterpart compliance empowerment
5.3 5.7 5.7
4.1 1.6 2.3
harmonization 5.8 traceability 6.1 literacy 6.3
3.0 0.6 1.1
ofusaitoseNtaa
?
-~
~-~
Recommended replacement
doorokakiN :lt.~·~{& iuumiNsaNkaku i:tf.\!;~'jiD kigyoosieN ~!::.t~ ho<-'>SiQkoo *'fAH kyoodookigyootai :Ji,: :rH~:t1¥ geNsiryokuboosaiseNtaa JJf.. -T- j; ~ff y~ ·{?: ~/ ~····riyoooiyas usa .f~ )'f'J L 4:'='""9 ~ secumeisekiniN l;:)t tYl" *11keNtookadai #t%rl·~;lll! isokukaNja :fH']J.~:;ff tokubecusagyoohaN
*»li'F:l-*
0
haamonaiz.eesoN
p toreesabiri:tii '1 rite ram
taiooaite ::lt};((*l'JfhooreijuNsu */}"il '
'Wt JF-t ~ ffU7 kacuyoonooryoku
,. s
)'iL!fU~}J
saabeeraNsu sapuraisaido zeroemiqsoN
surveillance 6.4 supply side 6.6 zero em is sio ns 6.7
3.8 4.8 3.0
coosakaNsi~~J[.~
kyookyuugawa ~Mt~ haisucuzero ::I'IHl:Hin
Wbile still debating recommended replacements, between publication of the second (NINJAL 2003b) and third reports (NINJAL 2004c), the Loanword Committee was already surveying local government administrative bodies on the results published so far. A survey (NINJAL 2004b) ""·as conducted in late 2003 and administered to staff at 680 randomly selected local governments nationwide. Respondents were divided into f()ur groups: (A) local government head.s, (B) staff
203
204
Loanwords in Japanese Table6.3. Comparison oflength between the 20 lea&t comprehended loanwords and the Loanword Committee's recommended replacements (adapted from NINJAL 2006c, 2007) '1-stat·' loan
a b
d
e
# Chat·act.ers
Recommended replacement
roodopuraiSiNgu doorokakiN U *Rit paburiQkuiNbor- juumiNsaNkaku 1t ~*@ ubumeNto iNkyubeesoN kigyoosieN ~~)!::~ eNfoosume~1:o hoosiqkoo :Jit\1"f koNsoosiamu kyoodookigyootai
Loan
Replacement
#Moras
Loan
Replacement
9
4
9
6
1.3
4
13
8
7
4
8
3
8
6
7
5
7
9
9
9
9
13
8
6
8
7 8
6
# ~J iL~:$:1*. f
ofusaitoseNtaa
g
akusesibiritii
geNsiryokuboosaiseNtaa Jrj-'f ~ ji~ ·t >/ ;)
.n
,Jfi!]}'f'Jl/-'?>~~
h j
akauNtabiritii ajeNda resipieNtO tasukufoosu
secumeisekiniN "IDi. W~ ~}(keJ>.1:ookadai titM~~ iSokukaNja if!i;tif..~~ tokubecusagyoohaN
9
4
9
4
4
4
7
6
4
6
6
6
5
6
9
8
4
8
8
4
8
8
4
8
7 7 7 4
4tJ~1i":t'* m n 0
p q
taiooaite .1.f }i~ it!* hooreijuNsu ~i-4--:l!!~ nooryokukaika ff~}J ll!l1t keNgeNfuyo ~ ll11-!oJ haamonaizeesoN kyoocoo ·~ ~~ toreesabiritii rirekikaNri ).l~~:f!Jl riteraSii yomikakinooryoku kauNtaapaato koNpuraiaNsu eNpawaameNto
6
4 9
2
9
8
4
8
5
6
8
4
8
6
WtJ:i.~~i:.:I.J
kacuyoonooryoku ;wm~m,7
r
t
saabeeraNsu sapuraisaido zeroemiQSON
coosakaNsi til! 1:1\:tl'!. kyookyuugawa *~1!!! haiSucuzero ~~ lli .CP
7 7 7
AVERAGE
7.7
4
6
3
6
4
6
4.3
7.7
7.1
responsible for printed information, (C) staff responsible for homepage mainte .. nance and (D) ordinary staff in direct contact with the local community. One per-son from groups (A), (B) and (C) was interviewed trom each of the offices, while 33 people from group (D) were interviewed from each. A range of questions was posed regarding the recommended replacements, as well as the role and existence of the Loanword Committee itself. I consider some of the more important below. While the majority of the group (A) and (B) respondents ( 60% and 53%, respectively) were aware of the Committee's reports, this was not the case with
Chapter 6. Attitudes to loanwords
the other two groups. Some 68% of group (D) staff were unaware, especially younger staff and those working in accounts and finance departments. Significantly, 70% of those 'Norking in welfare and 74% of those 'Norking in health and insurance departments, two domains especially singled out for low levels of loanword comprehension (NINJAL 2004a), were unaware of any loanword replacement proposals. When questioned about the necessity of replacements, however, more than two--thirds of all four groups believed they were required. 1he survey also posed a number of questions about specific replacements published in the Committee's first h\'o reports (NINJAL 2003a, 2003b ). I take up h\'o of these in Table 6.4, which cites percentage figures for group (D) respondents only (n 14,217). Table 6.4. Attitude oflocal government workers to Loan Committee replacement proposals (adapted from NINJAL 2004b, 2007) (i) Of tlze following pairs, which is easier to understand? The lo.mword or t1ze proposed non··loan repl.~eement?
Loan is easier
Loan replacement is easier
Impossible to say
Don't
know
Level of comprehension (NI NT AL 2007)
20%
,!9%
26%
1%
'H%
16%
44%
33%
4%
23%
41%
.33%
23%
I%
77%
(ii) The following· lotm wortls represent ideas not prel•iously expre.'i.~ible in Japanese. Do you think we should keep using these words?
Yes
No
Impossible to say
No answer
Level of comprehension (NINTAL 2007)
iNfoom udokoNseNto 'informed
15%
58%
25%
2%
23%
g-uroobaru/Cikyuukibo ~1~;1:)'/i~
'global'
iNfoomudokoNse.Nto/ naQtokusiNSaCU lit'J1fl. !$-.~
'informed consent' deesaabisu!higaerikaigo
S ~~ ;) 1Hl: 'ambulatory care'
ronsent' noomaraizee.{oN 'normalization' simyureeSoN 'simulation' bariafurii 'disabled access'
19%
54%
25%
2%
12%
51%
23%
24%
2%
58%
69%
14%
16%
2%
73%
Judging from the responses to question (i) in Table 6.4, reaction to the Committee's replacement recommendations was favourable. General levels of comprehension for both the loans guroobaru and iNfoomudokoNseNto were low and local government workers found the non·loan replacement easier to understand. This \·\'as not the case for higaerikaigo, the replacement suggested for deesaabisu, a loan which showed a 77% comprehension rate in NINJAL (2007: 265-289). Question (ii) in Table 6.4 sho'NS that a positive attitude to'~Nards the continued use of a loan is
205
206
Loanwords in Japanese
closely correlated with its level of comprehension. Both these questions had been posed of the general public a few months earlier in NINJAL (2(X)4a) . .Response pat·· terns were broadly similar, although the proportion of'impossible to say' answers \·\'as considerably lower, the higher reluctance to answer evident from Table 6.4 perhaps reflecting the conservatism of the Japanese, or any, civil service. Yamada (2005: 210···224) is particularly critical of the Loanword Committee's entire raison detre, regarding replacement proposals as 'useless' (opus cit.: 222) and likening them to the failed effi)rts of the Academic fran~aise at restricting English loans in French . .Replacements focus purely on the written and ignore the spoken word. Although writers may not always fully understand the meaning and their choice may not have been based on reasons of clarity, they typically have reasons for choosing a loam\'ord such as eNpawaameNto and responsibility lies with them. Yamada finds it difficult to imagine anyone leafing through any future 'loanword replacement manual' while penning an essay or newspaper article. If such a hypothetical manual had any use he believes it would only be for post--production edit·· ing purposes. The reaction of the general public to replacements has been mixed. While the governmenfs eftorts towards engendering smoother communication between the ordinary citizen and the bureaucracy have been praised, there is also a recog·· nition that many of the recommended replacements are themselves difficult to understand. In some cases a recommended replacement fails to convey the full meaning of the loan. Th.e Committee has admitted (Tanaka 2005: 16; NINJAL 2006c: 210···211) that it received considerable criticism for its recommendation of naQtokusiNsacu *~ 1!H~ ~, lit. 'consented medical examinatiorl, tor iNjoomudokoNseNto 'informed consent'. No notion of any 'informing' on the part of the physician is contained in the Sino--Japanese replacement. In the end, the unfamiliarity of any new term, be it loan or non-loan, will engender some lack of comprehension.
References
First Attestation Sources
0
+
Arakawa Sobei (~ ~,f;~b-f~B·'Z7..), ed. 1977. Gairaigo jiten dai2 ban (Loanword Dictionary 2nd Edition). Tokyo: Kadokawa Shoten. NKD = ShOgakukan (;]'4~), ed. 2000--2002. Nihon kokugo daijiten (The Great Dictionary ofthe Japanese Language). Tokyo: Shogakukan. Saito Shizuka (~AU¥). 1967. Nihongo r1i c~voboshita orandago no eikyo (The Influence of Dutch on Japanese). Tokyo: Shinozaki Shorin.
Other References = Ateji Gairaigo Jiten Henshtl Iinkai (;~ t: !1r ;!jt"~Hif ~) (Kanji- Phonetic Loanword Dictionary Editorial Committee). 1979. Atejl gairaigo jlten (Kanji-Phonetic Loanword Dictionary). Tokyo: Kashiwa Shobo. Akamatsu, Tsutomu. 2000. Japanese Phonology: A Functional Approach. Munich: Lincom. Alderete, John. 1999. Morphologically Governed Accent in Optimality Theory. Ph.D. dissertation, University of Massachusetts. Aoki, Paul Ken. 1981. Gemination in Japanese. Ph.D. dissertation, University of Washington. AraiHakuseki(mr# ;;p). [1715] 1968. Seiyokibrm (RecordsoftheOccident). Tokyo: Heibonsha. Arai Masako ()1'(;#f.f_-'f·). 1993. Furansugo kara no shakuyogo ni mirareru sokuon choon (Soku-on et choo-on dans les mots empruntes au fram;ais). ~voto Sangyo Daigaku Kokusaigengo Gakka Ker1kyiijo JohO 15: 75-96. Arai Ma~ako & Kawagoe Itsue (}Jifr!l" \---:_) 7" ). 1996. ShakuyCigo no sokuon: Nansensugo ni yoru chikaku)ikken hokoku (Consonant gemination in loanwords: A perceptual experiment with nonsense word~)- Ms. ArakawaSobei (;!l't)ll J~Ji;-~r). 1931. Nihor1go to natta eigo (JapanizedEnglish). 'lbkyo: Kenkyo.sha. Atkinson, Hoffman. 1879. Revised and Bniarged Editi<m ofR-.:ercises in the Y<Jkohama Dialect. Yokohama: Japan Gazette Office. Baba Sadayoshi (,~J~~ cll) a.k.a. Baba Sajuro (.~~'Ji+ !ill), trans. 1813. Rogobunpo kilum (A Standard Russian Grammar). Edo: Sei.doka Bunko, Otstiki Bunko. Backhaus, Peter. 2007. Linguistic Landscapes: .A Comparative Study of Urban Multilir1gualism in 'lhkyo. Clevedon: Multilingual Matters. Backhouse, A. E. 1993. The Japanese Language. Melbourne: 0 UP. Bailey, Don CHfford. 1960. The Rakuyoshii. Monumenta Nipponiaz 16: 289-376. BBK = Bunkacho Bunkabu Kokugoka (::Ut!T :t(-t·t$ i!J iifr·,Jl!O (Agency for Cultural Affairs, Culture Bureau. National Language Division). 1995. KiJkugo ni kansuru yoronchosa heisei 7-nen 4-gatsu (Opinion Poll on the National Language, April 1995). 'lbkyo: Okurasho Insatsulcyoku.
AG]HI
s
208
Loanwords in Japanese BBK. 1997. Kokugo ni kansuru yoronchtJsa heisei 9-netl 1-gatsu (Opinion Poll on the National Language, January 1997). Tokyo: Okurasho In~atsukyoku. BBK. 1998. Kokugo ni kansuru yoronchosa heisei 9-nen 11/12-gatsu (Opinion Poll on the National Language, November-December 1997). Tokyo: Okurasho In~atsukyoku. BBK. 1999. K<Jkugo ni kansunt yoronchOsa heisei 11-nen 1-gatsu (Opinion Poll on the National Language, January 1999). Tokyo: Okurasho In~atsukyoku. BBK. 2000. K<Jkugo ni kansuru yoronchOsa heisei 12-nen 1-gatsu (Opinion Poll on the National Language, January 2000). Tokyo: Okurasho Insatsukyoku. BBK. 2001. Kokugo ni kansuru yorond1osa heL~ei 13--nen 1--gatsu (Opinion Poll on the National Language, January 2001). Tokyo: Okurasho Insatsukyoku. BBK. 2002. Kokugo ni kansuru yorond1osa heL~ei 14--nen 1--gatsu (Opinion Poll on the National Language, January 2002). Tokyo: Okurasho Insatsukyoku. BBK. 2003. Kokugo ni kansuru yorot1ch6sa heisei 14-nen 11/12--gatsu (Opinion Poll on the National Language, November-December 2002). Tokyo: Okurasho Insatsukyoku. BBK. 2008. Kokugo ni kansuru yorond1osa heL~ei 20--nen 3--gatsu (Opinion Poll on the National Language, March 2008). Tokyo: Gyosei. Beckwith, Christopher. 200-1. Koguryo: The Language oj"Japan's Continental Relatives. Leiden: Brill. Benson, John & Matsumura, Takao. 2001. Japan 1868···1945: From Isolation to Occupation. Harlow: Pearson Education. Bloch, Bernard. 1950. Studies in colloquial Japanese, IV: Phonemics. Language 26: 86···125. Bloomfield, Leonard. 1933. Language. New York NY: Holt, Rinehart and Winston Boxer, Charles. 1951. The Christian Century in Iapatl. Berkeley CA: University of California Press. Bunkacho (;l:.f.t!T) (Agency for Cultural Affairs). 1991. Gairaigo 110 hyoki (Loanword Orthogra-· phy). (http://www. bunka.go.jp/kokugo/main.asp?tl=list&id= 1000003933&clc= 1000000068) (11 June, 2010). Bynon, Theodora. 1977. Historical Linguistics. Cambridge: CUP. Campbell, Lyle. 2004. Historical Linguistics: An Introduction. Edinburgh: EUP. Carr, Den:l'.el. 1951. Comparative treatment of epenthetic and paragogic vowels in English loan words in Japanese and Hawaiian. University of California Publications in Semitic Philology 11: 13-26. Chin Rikiei (l).lefrft). 2007. Waseikango (Kango Made in Japan). In Nihongogaku kenkyu jiten (Ihe Research Encyclopaedia of Japanese Linguistics), Hida Yoshifumi (~ IJ:J ~;t), Endo Yoshihide OUiHl~), Kato Masanobu (;.T,J~iE1~"), Sato Takeyoshi (~&~~~), Hachiya Kiyoto (~-1'.'-m )-..) & Maeda 'Ibmiyoshi (]fB":J ·;t~) (eds), 150-151. Tokyo: MeijiShoin. Clankie, Shm.vn. 1999. On Brand Name Change: A Theory of Genericization. Ph.D. dissertation, University of Hawaii. Cooper, ivlichacl. 1974. Rodrigues the Interpreter: .t\11 Barly Jes-..1it in Japan and China. New York NY: Weatherhill. Cooper, ivlichael. 1976. The Nippo Jis.iw. Monumenta Nipponica 31: 417-430. Coulmas, Florian. 1989. The function ofloan translations for the modernization of]apanese during the Meiji period. 1ransactions of the Asiatic Socie~v 4: 49-66. Coveney, Aidan. 2001. Ihe Sounds of Contemporary Frendt: .t\rticulation and Diversi~v. Exeter: Elm Bank Publications. Crawford, Clifford. 2007. The role of loanword diffusion in changing adaptation patterns: A study of coronal stops in Japanese borrowings. Working Papers of the Cornell Phonetics Laboratory 16: 32···56.
References Crawford, Clifford 2008. An evolutionary account of loanword-induced sound change in Japanese. University of Pennsylvania Workir1g Papers in 1JnguL~tics 14: 57-70. Dalkokuya Kodayu (;kl1Vi-Jt;&:.jJ. [1793] 2003. Hokusa ibun (Strange 'Hues from the North). In
Daikokuya Kodayu siryoshu, dai-2 kan. f(voryu to hyohaku no june,!: areuto rettlJ kara shiberia e, scshlf:t: peteruburugu ('Ihe Collected Historical Archives ofDaikokuya Kodayii, VoL 2: Ten Years of Drifting and Roaming: From the Aleutl.ans to Siberia and on to St. Petersburg), Yamashita 'l:'suneo (ill "f tOd (ed), 211-322. 'Ibkyo: Nihon Hyoronsha. Daniels, F.J. 1948. The vocabulary of the Japanese ports lingo. Bulletin of the School of Oriental and African Studies 12: 805-823. Daulton. Frank. 2008.Japan:S Built-In Lexicon ofEnglish-Based Loanwords. Clevedon: Multilingual Matters. DeChicchis, Joseph. 1995. The cw·rent state of the Ainu language. In Multilingual Japan, John Maher & Kyoko Yashiro (eds), 103---124. Clevedon: Multilingual Matters. Dohlus, Katrin. 2005a. Phonetics or phonology: Asymmetries in loanword adaptations French and German mid front rounded vowels in Japanese. ZAS Papers in Linguistics 42:117-135. Dohlus, Katrin. 2005b. On the asymmetry in the adaptation pattern of German umlaut in Japanese. Onsei Kenkyu (Tournai of the Phonetic Society ofJapan) 9(1): 39-49. Doi Tadao (1:}f,!!1. 4:.). 1933. Nihon yasokai noyogo ni tsuite (On the Terminology of the Jesuits in Japan). Gairaigo Kenkyu 3: 7-· 22. Doi Tadao. 1963. Kirishitan but~h~n ko (Kirishitan Literature). Tokyo: Sanseido. Doi Tadao. 1971. I(lrishitangogaku ,10 kenkyu (The Linguistic Kirishitan Materials). Tokyo: Sanseido. Doi Tadao, Takeshi Morita ~- li:l iii:.) & ChOnan Minoru (·J.- ii ~) (eds). [ 1603] 1980. Vocabulario da Li,zgoa de Japam com AdeciaraftlO em Portugues, e Irmaos da Companhia de Iesu. Tokyo: Iwanami Shoten. Doi, 1bshlo. 1976. :The Study ofLanguage in Japan. 'Ibkyo: Shinozaki Shorin. Downes, William. 1984. Language and Society. London: Fontana. Dudenreaktion (eds). 1993-1995. Duden: Das Grofte W6rterbuch der Deutschen Sprache. Mannheim: Dudenverlag. Dupoux, Emmanuel, Kakehi Kazuhiko, Hlrose Yuki, Pallier, Christophe & Mehler, Jacques. 1999. Hpenthetl.c vowels ln Japanese: A perceptual illusion? Journal of Experimental Psychology: Human Perception and Performance 25: 1568-1578. Ebisawa Arimichl. (~~-R1f1f). l99l. Klrishitan nanban bungaku nyumon (Introduction to Nanban Kir!sh!tan Literature). Tokyo: Kyobunkan. ETS Educational Testing Service. 2009. Test and Score Data Summary for TOEFL InternetBased and Paper-Based Tests: January 2009- December 2009 Test Data. (http://www.ets. org/Media!Tests/TOHFL/pdfitest_score_data_summary_2009.pdf) (11 January, 2011). Fabian, Fukansai [1592] 1967. Amakusabon Helke Monogatari (Feiqe no Monogatari (Tales of the Helke) Ama:ku.~a Copy). 'Ibkyo: Ofu.~ha. l1ujimoto Hlkaru (Jl;,t.;\t). 1993. Gaikoku chime!. jinmei no hyoki (The orthography of foreign toponyms and anthroponyms ). Gengo Seikatsu 266: 65-72. Fukuzawa, Haruka & Kitahara, Mafuyu. 2005. Ranking paradoxes in consonant voicing in Japanese. In Voicing in Japanese, Jeroen van de Weijer, Kensuke Nanjo & Tetsuo Nishihara (ed~), 105-121. Berlin: Mouton de Gruyter. Fukuzawa, Haruka, Kl.tahara, Mafuyu & Ota, Mltsuhiko. 1998. Lexical stratification and r..uli
209
210
Loanwords in Japanese Derrick Hlggins, Kenneth Olson & Tamra Wysocki (eds), 47-62. Chicago IL: Chicago Linguistic Society. Fuku1.awa Yuki.chi (;f.~~~h~' ). 1860. ZiJtei kaei tsugo (Engllsh-Japanese Dlctionary). Edo: Kaldo ?..ohan. Fukuzawa Yuki.chl. 1899. Fukuo jiden (Autobiography of Fukuzawa YU.klchi). Tokyo: Jijishinbosha. Gabbrielli, Richard. 2005 . .A Sociolinguistic .Analysis ofF:ngli.
References Heinrkh, Patrick. 2004. Language planning and language ideology in the Ryiikyu Islands. Language Policy 3: 153-179. Hepburn, James (ed.). 1867. A Japanese and English Dictionary/Waei gorinshusei. Shanghai: American Presbyterian Mission Press. Hiblya, Junko. 1995. 'Ihe velar na.~al in 1bkyo Japanese: A case of diffusion from above. Language Variation and Chattge 7: 139-152. Hiblya, Junko. 1996. De nasalization of the velar nasal ln 1bkyo Japanese: Observations in real time. In Towards a Social Science of Language. Papers in Honor of William Labov, Vol. I: Variation and Change in Language and Society, Gregory Guy, Crawford Feagin, Deborah Schiffrin & John Baugh (eds), 161-171. Amsterdam: John Benjamins. Hibiya, Junko. 1999. Variationist sociolinguistics. In The Handbook of Japanese Linguistics, Natsuko Tsujimura (ed.), 101-120. Oxford: Blackwell. Hida Yoshifumi (~ re ~Lx). 1977. Roshiagaktt shiryo (Russian study materials). In Koku-gogaku kenkyr4jiten (Research Dictionary of the National Language), Sato Kiyoji (tf.J.(!(. ~), 781-783. Tokyo: Meiji Shoin. Hida Yoshifumi 1998. Gairaigo no toreirikata no henka (Changes in how loanwords are borrowed). Nihongogaku 17(6): 29-38. Higashibaba, Ikuo. 2001. Christianity In Early Modern Japan: Kirishitan Belief and Practice. Leiden: Brill. Higashibaba, Ikuo. ~ ~.l;!;ff, S:.). 2006. Kirishitanshi salko: shinkojuyo no shukyogaku (Kirishitan History Reconsidered: Religious Studies on the Reception of a Faith). Tenri: Tenri Daigaku Fttzoku Oyasato Kenkyujo. Hirata, Yukari. 1990. Perception of geminated stops in Japanese word and sentence levels. Onsel Gakkal KalhO (Bulletin of the Photzetic Society ofJapan) 194: 23···28. Hirayama, Manami. 2008. Asymmetries in voiced stop gemination. In Asymmetries in Phonology: An East-Asian Perspective, Haruo Kubozono (ed.), 77--92. Tokyo: Kurosio. Hiro7.ane, Yoshito. 1992. Percepti.on by Japanese speakers of some English sounds as the Japanese choked sound iQi. Onsei Gakkai Kaihi:i (Bulletin of the Phonetic Society of Japan) 201:15-19. Hisamatsu Sen'ichi (;'z;f!lt}-) & Sato Kei1ZO (fr:JIUJ:.~.) (ed.). 1969. Kokugo }iten Oapanese Dictionary). Tokyo: Kadokawa Shoten. HNKK Homusho Nyukoku Kanri Kyoku (:),!,;~~A I@ %Wi M) Oapanese Ministry of Justice, Immigration Bureau). 2009. Kokuseki (shushinchi) betsu gaikoku}in torokushasu no suli (frend.~ on Registered Aliens, Analysed by Nationality or Birthplace). (http:/ iwww.moj. go.jp/content/000008047.pdf) (12 January, 2011). Hock, Hans. 1991. PrinciplesofHistoric.ai Linguistics. Berlin: Mouton de Gruyter. Holm, John. 1988. Pidgins and Creoles, Vol. 1: Theory and Structure. Cambridge: CUP. Holm, John. 1989. Pidgins and Creoles, Vol. 2: Reference Survey. Cambridge: CUP. Honna, Nobuyuki. 1995. English in Japanese society: Language within language. Journal of Multilingual and Multicult-ural Development 16: 45-62. I 'lbkubon (<$": ~._ ). 1988. Kankoku jinmei/chime! no nihongo kanahyoki ni tsuite ('[he Japanese Kilna transcription of Korean anthroponyms and toponym.~). Nihongogaku 10.7: 43-54. Ichikawa, Sanki. 1929. Foreign Influences In the Japanese Language. Tokyo: Japanese Council, Institute of Pacific Relations. Ichikawa, Sanki. 1930. 1he pronunci.ationofEnglL~h loan-words in Japanese. In A Grammatlr.al Miscellany Offered to Otto Jespersen on His Seventieth llirthday, N. B0gholm, Aage Brusendorti & C.A. Bodelsen (eds), 179···190. Copenhagen: Levin & Munksgaard.
21.1
212
Loanwords in Japanese Ide Junko (#-f-fi~-i· ). 2005. Gaikoku chime'! hyok'r n'r.tsuite (Forelgn placename orthography). In Hokoku 122: Zas.shi "laiyo' ni yoru kakuritsuki gendaigo no kenkyii (Repo.rt #122: Research on the Formative Era of Contemporary Japanese Based on the Taiyi) Corpus), Kokurltsu Kokugo Kenkyujo (~I ir.~l1#·f1fJi:IJj-) (Natlonallnstltute for the Japanese Language) (ed.), 157-172. 'Ibkyo: Hakubunkansh'rnsha. Inamura Sanpal
References Masanobu (ff,11fiEft), Sato Takeyoshi UillJi\~), Hachlya Kiyoto (~,:;HtJ...) & Maeda 'Ibmiyoshl ("iJ!r W!H}l:) (eds), 382-383. Tokyo: Melji. Shoin. Ito, Junko. 1990. Prosodic MinimaHty in Japanese. In Papers from the 26th Regional Meetir~g of the Chicago Linguistics Society, Vol. 2: 1he Parasession on the Syllable irt Phonetics and Phonology, Michael Zi.olkowski, Manuela Noske & Karen Deaton (eds), 213-239. Chicago IL: Chicago Linguisti.cs Sodety. Ito, Junko, Kitagawa, Yoshlhisa & Mester, Armin. 1996. Prosodic faithfulness & correspondence: Evidence fi"om a Japanese argot. Journal ofEast Asian Linguistics 5: 217--294. It6, Junko & Mester, Armin. 1986. The phonology of voicing in Japanese: Theoretical conse-quences for morphological accessibility. Linguistic Inquiry 17: 49--73. It6, Junko & Mester, Armin. 1992. Weak Layeritlg and Wi.1rd Binarity. Report LRC--92-09, Uni-versity of California, Santa Cruz. Republished (2003) in A New Century of Phonology at1d Phonological Theory: A Festschrift for Professor Shosuke Haraguchi on the Occasion of his Sixtieth Birthday, Takem Honma, Masao Okazaki, Toshiyuki Tabata & Shin-ichi Tanaka (eds), 26--65. Tokyo: Kaitakusha It6, Junko & Mester, Armin. 1995. Japanese Phonolog}' In The Handbook ofPhonoiogicai Theory, John Goldsmith (ed.), 817--838. Cambridge MA: Blackwell. It6, Junko & Mester, ArmiiL 1999. The phonologicallexicoiL In The Handbook of Japanese Linguistics, Natsuko Tsujimura (ed.), 62--100. Oxford: Black\vell. It6, Junko, Mester, Armin & Padgett, Jaye. 1995. Licensing and underspecification in optimality theory. Linguistic Inquiry 26: 571--613. Ito Masamitsu (·,-1UlJt). 2003. Kayokyoku no naka no gairaigo gaikokugo (Loanwords and Foreign Words in Pop Songs). Nihongogaku 22(7): 40---48. Janhunen, Juha & Salminen, Tapani. 1993--1999. UNESCO Red Book on Endangered Languages: Northeast Asia. (http:llw>vw.helsinki.fi/--tasalmin/nasia...report.html) (11 January, 2011). Jansen, Marius. 1984. Rangaku and westernization. Modern Asian Studies 18: 5·11---553. JET Japan Exchange and Teaching Programme. 2010. JET Programme Participant Numbers. (http://v.•ww.jetprogram.org/e/introduction/statistics.htrril) (13 January, 2011). Jinnouchi Masataka (M!Jkl IE.Jil<). 2000. 'Yakusho kotoba' no kai1.en manyuaru to 'yakusho katakanago' ('Municipal office language' improvement manuals and 'municipal office loanwords'). Nihongogaku 19(2): 62-72. Jinnouchi Masataka. 2007. Gairaigo 110 shakaigengogaku (The Sociollnguistics of Loanwords). Kyoto: Sekaishisosha. JKS Jiyukokuminsha ( i!J 1i! ~ R;~L) (l:ree Citizens Company) (ed.). 1960. Ge1~dai yogo no kiso chishiki 1960 (Essential Modern Vocabulary 1960). Tokyo: )iyiikokuminsha JKS (ed.). 1977. Gendai yogo no kiso dtishiki 1977 (Essential i•.,.fodern Vocabulary 1977). 'Jbkyo: Jiyiikokuminsha. JKS (ed.). 1978. Gendai yogo no kiso chishiki 1978 (Essential Modern Vocabulary 1978). 'Jbkyo: Jiyiikokuminsha. JKS (ed.). 1980. Gendai yogo no kiso chishiki 1980 (Essential Modern Vocabulary 1980). 'Jbkyo: Jiyiikokuminsha. JKS (ed.). 2003. Gendai yogo no kiso chishiki 2003 (Essential Modern Vocabulary 2003). 'Jbkyo: Jiyiikokuminsha. JKS (ed.). 2010. Gendai yogo no kiso chishiki 2010 (Essential Modern Vocabulary 2010). 'Jbkyo: Jiyiikokuminsha. Joo Hakutaro (:lli/,J:f_f!§';k~Jl). 1998. Nihongo onseikagaku (Japanese Phonetics). 'lbkyo: Bandai. Kabajima Tadao (~ ~ ~~ ;t:), ed. 1984. Meiji--Taisho shingo zokugo jiten (A Dictionary of Meiji and TaishO Era Neologisms and Colloquialisms). Tokyo: Tokyodo.
213
214
Loanwords in Japanese Kaiser, Stefan. 1991. Chinpanese? h'estern Loans itl _Japan, Bakumatsu-Afeiji. In Bumpeatl Studies itl Japanese Linguistics 1988-90, Lone Takeuchi (ed.), 49-62. London: Lone Publications. Kanagaki Robun ('~~ ;l:_[ I· _Y;:). 1871-72. i\guranabe (The Beefeater). 'lbkyo: Seishido. Kanagaki Robun. [1872-76] 1980. SeiyodochU hizakurige. "lbkyo: Senshiisha. Kanai Yoshimitsu (4:# IE ;it.). 1995. Gairaigo ni okeru sokuon (Moraic Obstruents in Japanese Loanwords). Onsei Gakkai KailuJ (Builetin of the Phonetic Society of_!apan) 208: 18-28. Kang, Yoonjung & Rice, Keren. 2008. Introduction. Journal ojBast Asian Litlguistics 17: 273--278. Kanno Ken (f J'f1f). 1985. Yogo no ryakugokei (The abbreviated forms ofJapanese loanwords). Nihongogaku 4(9): 54--64. Katayama, Motoko. 1995. Loanword accent and minimal reranking in Japanese. Phonology at Santa Cruz 4: 12---28. Katayama, Motoko. 1998. Optimality Theory and Japanese Loanword Phonology. Ph.D. disser-· tation, University of California at Santa Cruz, Kawagoe Itsue (!:L~l·"''::J.~.). 1995. Shakuyogo ni miru sokuonka to riz.umu shototsu (Conso-nant gemination in loanword phonology and rhyihm clash). Gengo Kenkyu 108: 46--73. Kawagoe Itsue & AraiMasako (~#$.·1- ). 2002. Shakuyogo ni okeru sokuon (Consonant gemi-· nation in loanwords). Ot1sei Kenkyu (Journal of the Phonetic Society ofJapan) 6(1 ): 53--66. Kawahara, Shigeto. 2008. Phonetic naturalness and unnaturalness in Japanese loanword pho-nology. Journal ofBast Asian Litlguistics 17: 317---330. Kindaichi Kyosuke (4: B:J ·-% rw) (ed.). 19-13. Meikai jiten (Concise Dictionary). Tokyo: Sanseido. Kindaichi Kyosuke (ed.). 2002. Shinsetl kokugo jitetl daihachiban (New Japanese Dictionary, 8th edn). Tokyo: Shogakukan. Kitahara Mafuy11 (~t 1~ J!.:.-t ). 1997. 0 n'imon to bunpo (Phonology and grammar). In Butpo to onsei (Grammar and Speech), Speech and Grammar Research Society C-*Jl
References 215 Kubo:wno, Haruo. 1999b. Mora and syllable. In 11u: Handbook of.Tapanese Llr1guistics, Natsuko 'Isujimura (ed.), 31-61. Oxford: Ulackwell. Kubozono, Haruo. 2001 a On the markedness of diphthongs. Kobe Papers ir1 Linguistics 3: 60-73. Kubozono, Haruo. 2001 b. Epenthetic vowels and accent in Japanese: Facts and paradoxes. In Issues in Jtiptmese Phonology and Morphology, Jeroen van de Weljer & Tetsuo Nishihara (eds), 113-142. Berll.n: Mouton de Gruyter. Kubozono, Haruo. 2002a Prosodic structure ofloanwords in Japanese: Syllable, structure, accent and morphology. Onsei Kenkyu Oournal of the Phonetic Society ofJapan) 6(1): 79-97. Kubozono, Haruo. 2002b. The syllable as a unit of prosodic organization in Japanese: Syllable structure, accent and morphology. In The Syllable In Optimality Theory, Caroline Fery & Ruben van de Vijver (eds ), 129-156. Cambridge: CUP. Kubozono, Haruo. 2006. Where does loanword prosody come from? A case study ofJapanese loanword accent. Lingua 116: 1140--1170. Kubozono, Haruo. 2007. Gairaigo no oiiinkozo (The phonological structure ofloanwords ). Gekkan Gengo 36(6): 60-67. Kubozono, Haruo & Fujiura, Yayol 2004. Morpheme-dependent nature of compound accent in Japanese: An analysis of'short' compounds. On'in Kenkyu (Phonological Studies) 7: 9-16. Kubozono, Haruo & FUkui, ~1isa. 2006. Phonological structure and unaccented nouns in Tokyo and Osaka Japanese. In Japanese ar1d Korean Linguistics, VoL 1-i, Timothy Vance & Kimberley Jones (eds), 39--50. StanfordCA: CSLI. Kubozono, Haruo & Mester, Mester. 1995. Foot and accent: New evidence from Japanese com-pound accentuation. Paper Presented at the Annual Linguistics Society ofAmerica Meeting, New Orleans. Kulmus, JohaniL 1722. Anatomlsche Tabellen (Anatomical Charts). Danzig. Kulmus, Johann. 1734 (trans. Gerardus Dicten, a.k.a Gerrit Dickten). Ontleedlwndige Tafelen (An..'ttomic."ll Charts). Amsterdam: de Janssoons van Waesberge. Kume, Kunltake, Healey. Graham & Tsuzuki, Chushichi. (eeL~). 2002. The Iwakura .Embassy, 1871-1873: A True Account of the r1mbassador Extraordinary and Plenipotentiary~ _Tourney ofObservation 1hrough the United States ofAmerica and Europe. London: Routledge. Kuwamoto Yuji (~:~.:::). 1998a Ryakugo kara mita nihongo no chojfionsetsu no kozo ni t.suite (Truncation and the Structure of Trtmoraic Syllables in Japanese). Gengo Kagaku Ronshu 2:25-35. Kuwamoto Yii.li. 1998b. Nihongo ni okeru fUkugogoryakugo no on'lnk6z0 (The phonologi.cal structure of Japanese clipped compound.~). On'in Kenkyu (Phonological Studies) 1: 161-168. Labrune, Laurence. 2002. The prosodic structure of simple abbreviated loanwords in Japanese: A constraint-based account. Onsei Kenkyu Uournal of the Phonetic Society of Japan) 6(1 ): 98-120. Labrune, Laurence. 2006. La Phonologie du ]aponais. Leuven: Peeters. LaCharite, Dadene & Paradis, Carole. 2005. Category prese.rvation and proximity versus phonetic approximation in loanword adaptation. Linguistic inquiry 36: 223-258. Ladefoged, Peter & Maddieson, Ian. 1996. 11-ze Sounds of the World~ Languages. Oxford: BlackwelL Lange, Roland. 1973. The Phonology of Eighth-Century Japanese. Tokyo: Sophia University Press. Lee, lksop & Ramsey, S. Robert. 2000. 1he Korean Language. Albany NY: State University of New York Press.
216
Loanwords in Japanese Long, Daniel. 2007. English on the Bonin (Ogasawara) L<Jar1ds. Durham NC: Duke University Press for the American Dialect Soci.ety. Loveday, Leo. 1996. Lar1guage Contact in Japan. ()}.ford: OUP. Lovins, Julie Beth. 1973. Loanwords and the Phonological Structure of)apanese. Ph.D. dissertation, University of Chicago. Lyman, Benjamin. 1894. Change from Surd to Sonant in Japanese compounds. Oriental Club of Philadelphia 1-17. MabuchiKazuo OUHtt;t:). 1993. Gojuonzu no hanashi (The Fifty Kana). Tokyo: Taishtlkan. Mackey, William. 1968. The description ofbilingualism. In Reading in the Sociology ojLanguage, Joshua Fishman (ed.), 554--584. The Hague: Mouton. McCawley, James. 1968. The Phonological Compotlent of a Grammar of Japanese. The Hague: Mouton. McCawley, James. 1977. Accent inJapanese. In Studies atid Stress and Accent, Larry Hyman (eel), 261--302. Los Angeles CA: University of Southern California Department of Linguistics. McCawley, James. 1978. \'\That is a tone language? In Tone: A Linguistic Surv<;_v, Victoria Fromkin (ed.), 113--131. New York NY: Academic Press. McConnell, Da'tid. 2000. Importing Diversity: Inside Japan's JET Program. Berkeley CA: University of California Press. Maeda Taro (mtrlfi ;f.dt~). 1922. Gairaigo no kenkya (Research on Loanwords). Tokyo: Iwanami Shoten. Maekawa, Kikuo & Hideaki, Kikuchi 2005. Corpus--based analysis of vowel devoicing in spontaneous Japanese: An interim report. In Voicitlg in Japanese, Jeroen van de Weijer, Kensuke Nanjo & Tetsuo Nishihara (eds), 205---228. Berlin: Mouton de Gruyte1: Maeno Ryotaku (ilt]gf ~.$!(). [1785] 1994. Oranda yakusen (Compendium of rules for the translation of Dutch). In Mae no rankashu (Towards Dutch Learning: An Anthology of the Works of Maeno Ryotaku), Sugimoto Tsutomu W;t.·-::; (~
Maruyama Takao Ctt.w:#))), Kobayashl Tadao (;J\;f*.,!t:?k:), Yama7..aki Chiakl. (w~~'fiU & 'Ierauchi Masanori ( <,f f,J i'E~). 1992. Katakanago o eigo ni suru jiten (Dlctionary of Katakana words and thelr English translations). Tokyo: Taishukan Shoten.
References Matsumori, Akiko. 1995. Ryiikyuan: Past, present and future. In Multilir1gual Japan, John Maher & Kyoko Yashiro (eds), 19-44. Glevedon: Multilingual Matters. Matsuoka Koji (.t£.·tlli;7tllj). 1982. Gairalgo no rekishi ('Ihe history of Japanese loanwords). In Goishi (Vocabulary History), Morioka Kenji. (~~1{.:::.) (ed.), 90-114. 'Ibkyo: MeijiShoin. Mester, Armi.n. 1990. Patterns oftruncation. Unguistic h1quiry 21: 478-485. Miller, Roy Andrew. 1967. '!he }apa11ese Language. Chicago IL: University of Chicago Press. Miller, Roy Andrew. 1998. Re..,iew of 'Language Contact in Japan: A Sociolinguistic History (Loveday 1996): journal ofJapanese Studies 24: 208-212. Miyajima Tatsuo ( t -it;;!<;). 1958. Kindai nihongo ni okeru tango no mondai (Some problems with modern Japanese words). Gengo Seikatsu 79: 17-26. Miyake, Marc Hideo. 1999. The Phonology of Eighth Century Japanese Revisited: Another Reconstruction Based Upon Written Records. Ph.D. dissertation, University of Hawaii. Miyake, Marc Hideo. 2003. Old Japatlese: A Phonetic Reconstruction. London: RoutledgeCurzon. Miyashita Akinobu ( #. r Be'f\f). 1993. Shappo to shapo: Furansugo shakuyogo ni okeru sokuon to choon ni tsuite (Shappo and Sfo.apO: /Q/-epenthesis and vowel lengthening in borrowings from French). Kyoto Sangyo Dalgaku Ronsha 24: 201-229. Mogami Katsuya (:f.J:.Jt~). 1984. Oyakusho wa katakanago ga osuki? (Do government agencies prefer loanwords?). NHK Hoso Kenkya to Chosa 34(8): 20-31. Mogi Toshinobu (_it*-~ fiji). 2007. Kokkai kaigiroku o mochiita gairaigo no bttnseki: 'inobeshon' o rei to shite (Usage of the loanword inobeshon in the Diet mintttes: An analysis). In Hokoku 126: Kokyobaitai no gairaigo galraigo iikae teian o sasaeru chosa kenkyu (Report 11126: Loanwords in the Public Media Basic Research on 'Paraphrasing Loanwords'), Kokuritsu Kokugo Kenkyiijo (OOir.OO~~~or) (National Institute for the Japanese Language) (ed.). 410···118. Tokyo: Kokw·itsu Kokugo Kenkyujo. MonbushO CY.iill1!) (Ministry of Education ofJapan). 1902. Gaikoku chimei jinmei torishirabe fukumeisho (Report on the Survey of Foreign Place and Personal Names). Japanese govern-ment archives cite report was read in committee 9 Aug. 1902. (ht1p:/IW¥.'W.bunka.go.jpi kokugo/file_l/1 0000 15351_2006nenpyo02.pdt) ( 14 June, 201 0). Monbusho. 1955. Gairalgo no hyoki (Loanword Orthography). 1bkyo: Meiji 1bsho. Moravcslk, Edith. 1978. Language contact. In Universals of Human Language, VoL I, Joseph Greenberg (ed), 93-122. Stanford CA: Stanford University Press. Mori, Arinorl. 1873. Bducation in Japan: A Series ofl,etters AddrP.ssed by Prominent Americans to Arinori Mori. New York NY: D. Appleton. Motoki Shoei (i-f..*-IE.*-) (ed.). [1814] 1976. Angeria gorintai.o;ei (General Collection of English Words). Tokyo: Yushodo. Motoori Norinaga (;.f..J!~-l). 1822. Kojiki den (Commentary on the Kojiki). Edo: Maegawa Rokuzaemon. Mugglestone, Lynda. 2003. Talking Proper: The Rise and Fall of the English Accent as a Social Symbol. Oxford: OUP. Murakami Hidetoshi Ctl..l:~~) (ed.) 1854. Sango benran (frilingual Vade Mecum). Edo: Yamashiroya Sahel. Murakami Hidetoshi (ed.) 1859. Furansu shirin (A Forest of French Word.~). (Publisher unknown). Murakami Hidetoshi (ed.) 1864. Futsugo meiyo (A Handbook of French). Edo: Tatsurido. Mutsukawa, Masah.lko. 2009. Japanese Loanword Phonology: The Nature of Inputs and the Loanword Suble.xicon. 1bkyo: Hitsuji Shobo.
217
218
Loanwords in Japanese Nagase Jiro (7)<.J*1W~ ~~). 1999. Go tw seisui: kyanpasukotoba no jumyo (Words waxing and waning: The Hfespan of campus words). Nihongogaku 18(9): 14-24. Nakagawa Yoshio ( 4' Jil ;t;it). 1966. Rendaku, rensei. (kasho) no keifu (A genealogy of sequential voicing and sequential non-voidng [Working Label]). Kokugo Kokubun 35: 302-314. Nakayama Erlko (o/ w~!fff), Kiryu Rika (;tliiJil:. iJ t.r•) & Yamaguchi Masaya (w ~ ~-@). 2007. Shinbun ni miru kikangairaigo (Core loanwords found in newspapers). In Hokoku .126: Kokyobaitai no gairaigo gairaigo iikae teiatz o sasaeru chosa kenkyu [Report #126: Loan-words in the Public lVIedia Basic Research on 'Paraphrasing Loanwords'], Kokuritsu Kokugo Kenkytljo ( ~1 ir. [ifH~·f!fj(; 1~) (National Institute tor the Japanese Language) (ed.), 343--378. Tokyo: Kokuritsu Kokugo Kenkyujo. Nakayama Yasuki ( 4' Ll;JjJl:-1jij-). 1997. Sama dezu: btchi bOizu ni sasagu (Summer Days: In Praise of The Beach Boys). Tokyo: Kosaido Shuppan. NHK = Nippon Hoso Kyokai ( 13 ;t.~l!ii$4:·) (ed.). 1987. NHK hOsa no kotoba gaidobukku (NHK Guide to Broadcasting \Nords). Tokyo: Nihon Hoso Shuppan Kyokai. NHK (ed.). 1998. Nihongo hatsuon akusento jiten (Dictionary ofJapanese Accent). Tokyo: Nihon Hoso Shuppan Kyokai. NHK (ed.). 2006. Shin hOsa gaidorain (New Broadcasting Guidelines). (http://www3.nhk.or.jp/ pr/keiei!bc-guideline/pdf/guideline.pdf) (13 January, 2011). NINJAL = Kokuritsu Kokugo Kenkytt)o (~] I£.1i!l1/i-·m%ffr) (National Institute for the Japanese Langttage, formerly The National Language Research Institute) (ed.). 1952. ShiryosM 2: Goi chOsa gendai shitzbun yogo no ichirei (Data Set #2: A Vocabulary Survey of Modern Newspapers). Tokyo: Shtlei Shuppan. NINJAL (ed.). 1962. Hokoku 21: Gendai zasshi 90-shu no yogoyc'iji daiichibunsatsu: hyc'iki oyobi goihyo (Report #21: Vocabulary and Chinese Characters in 90 Magazines of Today, Vol. I: General Description and Vocabulary Frequency Tables). Tokyo: Shiiei Shuppan. NINJAL (ed.). 1964. Hokoku 25: Gendai zasshi 90-shu no yogoyo}i daisanbunsatsu lli: bunseki (Report #25: Vocabulary and Chinese Characters in 90 Magazines of Today, Vol. lll: Analysis). Tokyo: Shuei Shuppan. NINJAL (ed.). 1970. Hokoku 38: Denslrikeisanki ni yoru slrinbun no goi chOsa II (Report #38: Computeri7.ed Survey ofNewspaper Vocabulary I I). Tokyo: Shuei Shuppan. NINJAL (ed.). 1983a. Hokoku 75: Gendai hyoki no yure (Report #75: Variability in Modern Orthography). Tokyo: Shuei Shuppan. NINJAL (ed.). 1983b. Hokoku 76: KokO kyokasho no goi ch6sa (Report #76: Studies on the Vocabulary of Senior High School 'Ib:tbooks). Tokyo: Shuei Shuppan. NINJAL (ed.). 1984. Hokoku 79: Zasslri .90-shu shiryo no gairaigohyoki (Report #79: Loanword Orthography in the '90 iv1agazines' iv1aterials). Tokyo: Shiiei Shupp an. NINJAL (ed.). 1986. Hokoku 86: Chugakko kyokasho no goi chOsa (Report #87: Studies on the Vocabulary ofJunior High School Textbooks). Tokyo: Shiiei Shuppan. NINJAL (ed.). 1987. Hokoku 89: Zasshi yogo no hensen (Report #89: Vocabulary Trends in Magazines). Tokyo: Shiiei Shuppan. NINJAL (ed.). 1990. Nihongo kyotku .Shido sanko.Sho #16: gairaigo 110 keisei to sono lcyOiku (Reference Guide for Japanese Language Educati.on #16: The Evolution of Loanword.s and 'Iheir Instruction). Tokyo: Ministry of Finance. NINJAL (ed.). 1999. Hokolcu 115: 1erebi hOso no goichOsa Ill: keiryotelci bunselci (Report #115: A Vocabulary Survey of Television Broadcasts III: Quantitative Analysis). Tokyo: Kokuritsu Kokugo Kenkytljo.
References NINJAL (ed.). 2003a. Dai ikkai gairaigo iikae teian: wakarirlikui gatraigo o wakartyasuku suru tame no kotobazukai no k~fU (Paraphrasing Loanwords: Making Difficult to Understand Loanwords More Intelligible. Proposal!.). 'Ibkyo: Kokuritsu Kokugo Kenkyujo. NINJAL (ed.). 2003b. Dai nikai gairaigo iikae teian: wakarinikui gairaigo o Wtikariyasuku suru flune r1o kotobazukai rw k~fa (Paraphra.~ing Loanwords: Making Difficult to Understand Loanword.~ More Intelligible. Proposal 2.). 'Ibkyo: Kokuritsu Kokugo Kenkyujo. NINJAL (ed.). 2004a. Gairaigo r1i kan.~uru ishikl chosa (A Survey on Attitudes to Loanwords). Tokyo: Kokuritsu Kokugo Kenkyujo. NINJAL (ed.). 2004b. Gyosel joiw o wakariyasuku tsutaeru kotobazukai no kuja ni kansuru ishikl chOsa Qichitai chOsa) (Survey on Attitudes to Loanwords Related to Making Difficult to Understand Administrative Information More Intelligible: Local Government Survey). Tokyo: Kokuritsu Kokugo Kenkyujo. NINJAL (ed. ). 200·ic. Dai sankai gairaigo iikae telan: wakarinikui gairaigo o wakariyasuku suru tame no kotobazukai no kuja (Paraphrasing Loanwords: Making Difficult to Understand Loanwords More Intelligible. Proposal3). Tokyo: Kokuritsu Kokugo Kenkyujo. NINJAL (ed.). 2005a. Gendal zasshi rw goi chOsa: 1994-nen hakkO 70-shi (A Vocabulary Survey of Modern Magazines: 70 Magazines Published in 1994). Tokyo: Kokuritsu Kokugo Kenkyujo. NINJAL (ed.). 2005b. Gairalgo ni kansuru ishiki chosa II (zenkokuchOsa) (A Survey on Attitudes to Loanwords II: The National Survey). Tokyo: Kokmitsu Kokugo Kenkyt1jo. NINJAL (ed.). 2006a. Gendai zasshi 2,000,000-ji gengo chOsa goihyo kokaiban, ver. 1.0 (Vocabu-lary Tables for the 2 Million Character Survey of Current Magazines: Public Version 1.0). (http:iiwww.kokkengo.jp/katsudoiseika/goityosa) (12 January, 2011 ). NINJAL (ed.). 2006b. Dal yorzkai gairaigo iikae teiarz: wakarirzikui gairalgo o wakariyas-uku suru tame rzo kotobazukai no kuja (Paraphrasing Loanwords: Making Difficult to Understand Loanwords More Intelligible. Proposal 4.). Tokyo: Kokuritsu Kokugo KenkyO..Jo. NINJAL (ed.). 2006c. Gairaigo iikae tebiki (A Guide to Paraphrasing Loanwords). 'Ibkyo: Gyosel. NINJAL (ed.). 2007. Hokoku 126: Kokyobaitai no gairaigo - gairaigo iikae teian o sasaeru chosa kenkyu (Report #126: Loanwords in the Public Media- Basic Research on 'Paraphrasing Loanwords'). 'lbkyo: Kokuritsu Kokugo Kenkyt1jo. Nishimura Hajl.me (l!§.f<j· f.-). 1995. Sabaibaru eigo no su:;ume (Towards Survival English). 'lbkyo: Chikuma Shinsho. Nishio 'lbraya (W~~'kjt). 2002. Goshu (Word types). In Gol lmi (Words and Meaning), Saito Michiaki (*Hfiit!9l) (ed.), 29-53. Tokyo: Asakura Shoten. Nomura Masaaki (Jt # .1ll!N ). 1984. Goshu to wgoryoku (Word classification and productivity). Nihongogaku 3(9): 40-54. Nomura Masaaki & Yanase Satoko (;~IIJ:~~'t ). 1979. Jido monogatari no goikow (The makeup of vocabulary in chUdre:ds stories). Keiryo Kokugogaku 12: 86-95. Nomura Masaaki & Yanase Satoko 1989. Statistical structure of vocabulary books for juveniles. In_Tapane.se Quantitative Linguistics, Shizuo Mizutani (ed.), 14-30. Bochum: Brockmeyer. Norman, Arthur. 1955. Bamboo English: The Japanese influence upon American speech in Japan American Speech 30: 44-48. NSK = NihonShikaku Kyokai (E! *-:t;l~~) Oapan Licensing Authority) (ed.). 1982. }lSkogyo yogo da~liten (The Great Di.ctionaryof)IS (Japan J:nd.ustrial Standard) Indu.~trial1erminol ogy). 'Ibkyo: Denpa Shinbunsha.
219
220
Loanwords in Japanese NSKH = Naikaku Soridaijln Kanbo Koh6shltsu (Pi ~~{l;JJ!!_;:*: E: '!! tfit'*-j£) (Cabinet and Prime Minister's Secretariat, Communications Office) (ed.). 1977. KiJkugo rzi kansuru yorom:hosa (Opinion Poll on the National Language, August 1977). 'Iokyo: Nalkaku Sorldaijln Kanbo Kohoshitsu. Numata, Jiro. 1964. The 'Introduction of Dutch language. Monumenta Nippon lea 19: 243-253. Numoto Katsuaki Orl?.ir.>i~~). 1986. Nihor1 kanjiort no rekishi (The Phonological Hi.story of Japanese Kanji). Tokyo: TOkyodo. Ogawa Kikumatsu (;j,Jil ~if'!}). 19-15. Nichibei kaiwa teeM (Notes on Conversing with Americans). Tokyo: Seibundo Shinkosha. Ohno, Kazutoshi. 2000. The lexical nature ofRendaku in Japanese. In ]apaneseiKorean Lin-guistics, Vol. 9, Nakayama Mineharu & Charles J. Quinn (eds), 151--16<1. Stanford CA: CSLI. Ohso, Mieko. 1971. A Phonological Study of Some English Loan Words in Japanese. MA thesis, Ohio State University. Ohye Saburo (k'li.~!l~). 1967. Gairaigochfl no sokuon ni kansum ichikosatsu (The mora phoneme /Q/ in English loan words inJapanese). Onsei no Kenkya 13: 111--121. Okaz.awa Hidetora ( fiUJ ~'{'flj Jit). 19<16. Roshiago h'iza (A Course in Russian). Tokyo: Jiji Tsflshinsha. Okuda, Kunio. 1975. Accentual Systems in the Iapa~tese Dialects. Hiroshima: Bunka Hyoron. Onishi Katsunari (Ki!Ji $:-1!!) & Kajiki Minako (;@~ ;;lc3t*-'f ). 1995a. Katakanago 'raifi.uain' wa wakarimasu ka? (Do you understand the meaning of the loanword 'raifi.train'?). Hoso Kntkya to ChOsa 45(9): 14--25. Onishi Katsunari & Kajiki Minako 0Jll:~;j--:.jlt*-·f). 1995b. Masukomi ni tsukawarete iru gai-raigo no jittai (Loanword Use in the Mass Media). Hoso Kenkya to ChOsa 45(10): 18---31. Ono Ko)i (;J\Jftti'li ifj). 1991. Gairaigo to shite no eigo no sokuonka ni tsuite (!qi--Insertion Rule in English Words in Japanese). Gengo Kr.~nkya 100: 67---88. Oshima ShOji (f.: ~.iL~-). 2006. ht Chinese Characters Denrai (The Introduction of Kan}i to Japan). Tokyo: Iwanami ShoteiL Osumi, Midori. 2001. Language and identity i.n Okinawa today. In Studies in }apaltese Bilinguaiism, Mary Goebel Noguchi & Sandra Fotos (eds), 68-97. Clevedon: Mtiltilingual Matters. Ota, Mitsuhiko. 2004. The learnability ofthe stratified phonologicallexicon.Iouma/ ofjapanese Linguistics 20: 19-40. ()taka, Hiromi. 2008. Phonetics and Phonology of Moras, Feet and Geminate Cmtsonants ill ]apa12ese. Lanham MD: University Press of Ameri.ca Otori, Ranzaburo. 1964. The acceptance of western medicine in Japan. 1\.i(mumenta Nipponica 19: 254-27 4. Otsu Yilkio. 1980. Some aspects of Rendaku in Japanese and related problems. MIT 1Vorking Papers in Linguistics, 2: Theoretical Issues in Japanese Linguistics, 207-227. Otsuki Fumihiko (~k~_Y:ff-). 1884. Gairaigogen ko (Notes on Loanword Etymologies). Gakugei Shirin 14: 122-139,370-384 & 572-590. Otsuki Fumihiko (ed.). 1889. Genkai (An Ocean ofWords). 'Ibkyo: Shfleisha. Otsuki Gentaku (;k_>W,§~_) (ed.). 1788. Rangaku kaitei (Ladder to Dutch Studies). Edo: Matsumoto Heisuke & Matsum.oto Zenbei. Pa.rk, Whaja. 1987. Western Loan-Words in Japanese. Stockholm: Stockholm University Department of Ori.ental Languages. Patil, Hermann. 1880. Prinzipien der Sprachgeschichte. Halle: Max Ni.emeyer.
References Penny, Ralph. 2000. Varlatior1 and Change in Spanish. Cambridge: CUP. Peperkamp, Shal'on, Vendelin. Inga & Kimihil'o, Nakamura. 2008. On the perceptual origin of loanword adaptations: Experimental evidence from Japanese. Plumology 25: 129-164. Polivanov, Evgenij. 1931. La perception des sons d'une langue ctrangere. Travaux du Cerde Linguistique de Prague 4: 79-96. Pon 1:ei (~~~). 2003. Nihongo no tokuc1:o (Idiosyncrasies of the Japanese Language). 'Jbkyo: Bonjlnsha Poser, William. 1990. Evidence for foot structure in .Japanese. Language 66:78-105. Preston, Dennis & Yamagata, Ayako 2004. Katakana representation of English loanwords: Mora conservation and variable learner strategies. Journal of Sociolinguistics 8: 359-379. Pulleyblank, Edwin. 1991. Lexicon ofReconstructed Pronunciation in Early Middle Chinese, Late Middle Chitlese, and Early Afandaritl. Vancouver: University of British Columbia Press. Quackenbush, Edward 1974. How Japanese borrows English words. Linguistics An Interna-tional Review 131: 59---75. Quackenbush, Hl.roko. 1977. English loanwords in Japanese: Why are they difficult for Englishspeaking students? The Journal of the Association of Teachers of Japanese 12: 149-173. Quackenbush, Hiroko (h'Y-7"~/·y-;/.::z.:lt-'f). 1989. Gairaigo no sokuonka ni tsuite (Gemination in Japanese loanwords). In Ryugakusei nihongokyOiku ni kansuru rirontekl jis..~entekl kenky1~ (Theoretical and Practical Research on Japanese Language Edu.. cation tor Foreign Students), Hiroshima Daigaku Kyol.kugakubtt (Jt. lift;k *""-~ (ed), 1... 7. Hiroshima: Hiroshima Daigaktt Kyoikugakubu Ryugakttseinihongokyol.ku! Nihongokyoiku Gakka. Rice, Keren. 1997. Japanese NC dusters and the redundancy of postnasal voicing. Linguistic Inquiry 28: 541 .. 551. Rice, Keren. 2006. On the patterning of voiced stops in loanwords in Japanese. Toronto Working Papers itl Linguistics 26: 11 ... 22. Rickmeyer, }ens. 1989. Sprachbau. In Sprad1e und Schrift Japans, Bruno Lewin (ed), 26-62. I.eiden: 13 rilL lJ~-if~) (Interim National Language Research Council). RKC = Rinji Kokugo Chosakai( ij; 1926. Kanazukai katteian hosoku: gaikokugo no utsushtkata (Supplement to the Proposal for the Revision of Kana Usage: The Transcription of Loanwords). (Publi~her unknown). Rodriguez,Joao. 1604. Arte da Lingoa deiapam. Nagasaki: Societyof]esus Mission Press. Romaine, Suzanne. 1989. Bilingualism. Oxford: Blackwell. Rosen, Eric Robert 2001 Phonological Processes Interacting with the Lexicon: Variable and Non-Regular Effects in Japanese Phonology. Ph.D. dissertation, University of British Columbia. Rothaug, Petra. 1991. Abrij.l der japanischen Lautgeschichte. Hamburg: Helmut Buske. Saglsaka, Y. & Tohkura, Y. 1984. Phoneme duration control for speech synthesis by rule. Trans. Inst. flectron. Commun. ]67-A.: 629-636. Saida 'lbshio (*m~~). Naba Daikichi Cll!l~~k-~), Kunishi Heiroku (~isPf:f.;) & Rudolf Lehmann (ed). 1877. Wadoku taiyaku jlrln/Worterbuch der japanischen und deutschen Sprache. Tokyo: Hlbiya Kenjl.ro and Kato Suikei. Saito 'Isuyoshi (:l-:&~. 1977. Meiji no kotoba: Higashi kara nishi e no kakebash! (Meiji Era Words: A Bridge From East to West). 'Ibkyo: Kodansha. Saito YO&hio (~fiUK~). 1997. Nihongo onseigaku nyumon (An Introduction to Japanese Phonetics). 1bkyo: Sanseido.
*i\1!)
*!@
221
222
Loanwords in Japanese
Saito Yoshlo. 2003. Gendal nihongo no onsei: Bunsetsuon to onseikigo (The sounds of Modern Japanese: Segmentati.on and symboli7.ation). In Onsei ort'in (Phonetics and Phonology), Kitahara Yasuo (At.~1lfd:fu) & Uwano Zendo U::J~r*~) (eds), 1-21. Tokyo: Asakura Shoten. Sakamoto Mltsuru (~f*Jt;). 2002. Do suru? Gairaigo no hyoki to hatsuon (How should loanwords be transcribed and pronounced?). Hosa Ki:nkyii to ChOsa 52(10): 50-71. Saneto Kelshu (;~~;~~). 1973. Kindai rtitchU koshosiwa (A History ofModern Sino-Japanese Negotiations). Tokyo: ShunsMsha. Satake Hideo (tf. #;%-;iii). 1981. Wago, nadtike, ryt1koka (Native vocabulary, naming and popu-lar songs). Gengo Seikatsu 359: 46---52. Satake Hideo. 1991. 'Shingenbun'ichitai' no keiryoteki bunseki ('New Genbunichi'-style quantative analysis). Aiukogawa Joshi Daigaku Gengobunka Ket1kyujo Nenpo 3: 1--14. Sawada Tazuko (~[:E) EB ~:'f). 1985. Gairaigo ni okeru bointenka ni tsuite (\Vhich Kana are used in Japanese loan--words borrowed hom words containing two-consonants series? (sic.)). Kokugogaku 1-13: 75--88. Schurhamme1; Georg. 1928. Das kirkliche Spradtproblem in der japanischen ]esuitenmis.~ion des 16. und 17. Jahrhunderts. Tokyo: Deutsche Gesellschaft flir Natur- und Volkerkunde Ostasiens. Seeley, Christopher. 2000. A History of Vlrititlg in lap an. Honolulu HI: University of Hawaii Press. Sekiguchi Tsugeo ( ~] p ~ JJ ). 1942. Hyojtm shoUl doitsugo ko.za jokan (A Standard Elementary Course in German, Vol. 1). Tokyo: Nikko Shoin. Sekiguchi Tsugeo. 1951. Yasashii doitsugo (Easy German). Tokyo: SansMsha. Sekine Ken'ichi (F) #f_ ~- ---). 2003. Shinbunkiji no naka no katakanago (Loanwords in newspaper articles). Nihongogaku 22.7: 30--39. Shibatani, Masayoshi. 1990. The Languages ofjapan. Cambridge: CUP. Shibu Shohei Ct *J! ll/F¥-) & Sanada Shinji (f!: ll:l~~~). 1980. Goi chosa to sono bunseki (A vocabUlary survey and its analysis). In Nihon}in no chishiki kaiso ni okeru hanashikotoba no jittai (The State of the Spoken Language Amongst Educated Japanese), Kokuritsu Kokugo Kenkyujo (00§~~-~~p.~-) (Nationa:t Institute for the Japanese Language) (ed.), 63-82. 'Ibkyo: Kokuritsu Kokugo Kenkyiijo Nihongo Kyoiku Senta. Shillony, Ben-Ami. l98l. Politics and Culture in Wartime Japan. Oxford: Clarendon Press. Shinmura Izuru (~t]'!L)) (ed.). 1955. Kojien dai 1 ban (Koji.en 1st edn). Tokyo: lwanami Shoten. Shinmura Izuru (ed.). 1969. K6jien dai 2 ban (Koji.en 2nd edn). 'Ibkyo: lwanami. Shoten. Shinmura Izuru (ed.). 2008. K6jien dai 6 ban (Koji.en 6th edn). Tokyo: Iwanami Shoten. Shinohara, Shigeko. 1997. Analyse Phonologique de fAdaptation Japonaise de Mots lit rangers. Ph.D. disse.rtation, Universite Paris III. Shinohara, Shigeko. 2000. Default accentuation and foot structure in Japanese: Evidence from Japanese adaptations of French words. journal of Bast .Asian Linguistics 9: 55-96. Shinohara, Shigeko. 2004. Emergence of Universa:t Grammar in foreign word adaptations. In Constraints in Phonological .Acquisition, Rene Kager, joe Pater & Wim Zonneveld (eds), 292-320. Cambridge: CUP. Shirafuji Noriytiki ( Hl:lf ¢ ). 2005. BunshO to kanji (Writing and Kanji). In Kanji to nihongo (Kanji and Japanese), Maeda Tomiyoshi (ilFt:FJ ;t;t,Jl:) & Nomura Masaaki (ft#~llfl) (eds), 207-218. 'Ibkyo: Asakura Shoten. Shirai, Setsuko. 1999. Gemination in Loans from English to Japanese. MA thesis, University of Washington.
s
References Sibata 'Iakesi (~ liJ ii:t,). 1994. Gairaigo ni okeru akusentokaku no ichi (The location of accent In loanwords).ln Gendai Mgen no kenkyii (Modern Dialect Studies), Sato K:iyoji (1£•t·'ft5M (ed.), 388-418. Tokyo: Meiji Shoin. Siddle, Richard. 1996. Race, Resistance and the Ainu o.f.Tapan. London: Routledge. Silverman. Daniel1992. Multiple scansions in loanword phonology: Evidence from Cantonese. P1wnoiogy 9: 289-328. Smith, Janet & Schmidt, David. 1996. Variability ln written Japanese: Towards a sociolinguistic of script choice. Visible Language 30(1): 46-71. Smith, Jennifer L. 2006. Loan phonology is not all perception: Evidence from Japanese loan doublets. In Japanese/Korean Li1lgr.tistics 14, Timothy Vance & Kimberly Jones (eds ), 63-7 4. Stanford CA: CSLI. Society of .Jesus (ed.). 1598. Racuyoxu (= Rakuyosha) (Collection of Fallen Leaves). Nagasaki: Society ofJesus Mission Press. Sago Masaaki (~~IEI!JJ) & Hida Yoshifi.tmi (~~tE jit_Y;:) (eds). 1986. Meiji no kotoba jiten (A Dictionary ofMeiji Era Words). Tokyo: TokyOdo. Sohn. Ho-min. 1997. Orthographic divergence in South and North Korea: Toward a unified spelling system. In The Koreatl Alphabet: Its History and Structure, Young-Key Kim-Renaud (ed.), 193-218. Honolulu HI: University of Hawaii Press. Someya Hiroko (~.~-m·T- ). 2002. Kanban no moji hyoki (Writing on signs). In Gendai nihot1go ko.za dai 6-ka11: moji hyoki (Modern Japanese Course, Vol. VI: Letters and Orthography), Hida Yoshifumi (~~ m~x) & Sato Takeyoshi ({{::_&iit,~) (eds), 221···243. Tokyo: Meiji Shoin. SSDH = Sanseido Henshujo (.:;:~ 1t~1H'D (Sanseido Editorial Institute) (ed.) 1979. Konsaisu gairaigo jiten dai 3 ban (Concise Loanword Dictionary, 3rd edn). Tokyo: Sanseido. SSDH (ed.). 1991. Konsalsu gairaigo jlten dai 4 ban (Concise Loanword Dictionary, 4th edn). Tokyo: Sanseido. SSDH (ed.). 2005. Deri kon.
22.3
224 Loanwords in Japanese Suzuki, I-Hsaml. 1995. Mlnlmal words in Japanese. Proceedings of the Chicago Linguistic-~ Society 31:448-463. SYK =Shinbun Y6go Kondankai (~jffl.!j)ifj~-~Ufi_-i>) (Newspaper Terminology Roundtable Conference) (ed.). 1984. Gairaigo no kakikata (\Vriting Loanwords). Tokyo: Nihon Shi.nhun Kyokai. Takagi, Naoytiki & Mann, Virginia. 1994. A perceptual basi.s for the systemati.c phonological correspondences between Japanese load (sic.) Words and their English Source words. Journal ofPhonetics 22: 343--356. Takashi, Kyoko. 1990a. A Functional Analysis of English Borrowings in Japanese Advertising: Linguistics and Sociolinguistic Perspectives. Ph.D. dissertation, Georgetown University. Takashi, Kyoko. 1990b. A sociolinguistic analysis of English borrowings in Japanese advertising texts. World Englishes 9: 327---341. Takayama Tomoaki ( ~ w~ ~~ ). 1999. Shakuyogo no rendaktt/koonka nits uite (Sequential voicing and fortis in loan vocabulary). Report of the Special Research Project for the Typological Investigation of Languages and Cultures of the East and West 1: 375--385. Takayama TomoakL 2005. A survey of Rendaku in loanwords. In Vt1icing in Japanese, Jeroen van de Wei)e1; Kenstike Nanjo & Tetsuo Nishihara (eds), 177--190. Berlin: Mouton de Gruyter. Tamaoka, Katsuo, Ihara, Mutsuko, Murata, Tadao & Lim, Hyunjung 2009. Effects of firstelement phonological length and etymological--type features on sequential voicing (Re1ldaku) of second elements. Jou1"1!al o!Japanese Linguistics 25: 17-- 38. Tamma Suzuko ( T:B ;f<j-·t-·:f-=f ). 1996. Ainugo saru hogu1 jiten (Ainu-Japanese Dictionary Sam Dialect). Tokyo: Softikan. Tanaka Makiro ( 11! lf' 1/J( !'!~ ). 2005. Gairaigo no ilkae to kihan (Loanword replacements and mod-els). Niho11gogaku 2,1.10:6---19. Tanaka Makiro. 2006. Gendai shakai ni okeru gairaigo no jittai (The actual state ofloanwords in modern society). In Gairaigo to gendai shakai (Loanwords and Modern Soclety), Kokuritsu Kokugo Kenkyiijo ( ~ § ~ ~+Jf~.f.~-) (National Institute for the Japanese Language) (ed), 38-47. Tokyo: Kokuritsu Insatstikyoku. Tanaka Maklro. 2007. Gairaigo no teichakudo deeta no bunsekl (Analysis ofloanword pervasion data). In llokoku 126: Kokyobaitai no gairaigo - gairaigo iikae teian o sasaeru chosa kenkyu (Report #126: Loanwords in the Public Media- Basic Research on 'Paraphrasing Loanwords'), Kokuritsu Kokugo Kenkyiijo (~it. I'm~*~~ lf,) (National Institute for the Japanese Language) (ed..), 311-322. Tokyo: Kokuritsu Kokugo Kenkyiijo. Tanaka Shin'ichL 1996. Nihongo no onsetsu ko:z.O to orrin genshO (Syllable Structure and Phonological Phenomena in Japanese). MA thesis, Osaka University of Foreign Studies. Tanaka Takehlko (Etllf''ft_ff'). 2002. Gairaigo to wa nani ka (What Are Loanwords?). Tokyo: Choeisha. Tanomura Tadaharu (EtJJTt)".!lJ.iA). 1999. Gairaigo akusento ni okeru gengo no hatsuon no kan'yo ni tsulte (On the relation of)apanese loamvord accentuation to the pronunciation of the original word). Nihongo Kagaku 5: 67-88. Tateishi, Koichi. 1989. 'Iheoretical implications ofJapanese musicians' language. In Proceedings of the Eighth West Coast Conference 011 Formal Ling-uistics, Jane Fee & Kathryn Hunt (eds), 384-398. Stanford CA: CSLI. Thomason, Sarah. 200 l. Language Contact: An Introduction. Washington DC: Georgetown University Press.
References Thomason, Sarah & Kaufman, Terrence. 1988. Language Cor1tact, Creolizatlon and Ger1etic Lir1gui.~tics. Berkeley CA: University of California Press. Tokieda Motoki (H*-'.tf..W#.:~) (ed.). 1956. Reikai kokugo jiten (An illustrative Dictionary of the Japanese Language). Tokyo: Chukyo Shuppan. 1bmoda, 'Iakako. 1999. The impact of loan-words on Modern Japanese. ]tiptm Forum 11: 231-253.
Toyama EiJi (:7!-U;~lk). 1972. Kindal. no on'ln (Phonology of the modern period). In Kc1za nihongoshl 2: on'inshi, mojishi (A Course in the History ofthe Japanese Language, VoL II: The History of Phonology and Orthography), Nakata Norio (If re .:fJ(.j;:) (ed.), 173-268. Tokyo: Taishii.kan Tranel, Bernard. 1987. The Sounds of French. Cambridge: CUP. Tranter, Nicolas. 2008. Nonconventional scrlpt choice in Japan International Journal of the Sociology ofLanguage 192: 133---151. Tranter, Nicolas. 2011. Graphic Loans: East Asia and Beyond. Word 58.1. Tranter, Nicolas. Forthcoming. Modern Japanese. In The Languages of Japan and Korea, Nic Tranter (ed.). London: Routledge. Trudgill, Peter. 2010. Investigations in SociohistoricalLinguistics: Stories ofColonisation and Contact. Cambridge: CUP. Tsubouchi ShOyo (:li'fkl~lt). 1885. Tosei shosei katagi (The Character of Modern Students). Tokyo: Banseido. Tsujimura, Natsuko. 2007. An Introduction to Japanese Linguistics. Oxford: Blackwell. Tsujimura, Natsuko & Davis, Stuart. 1987. The accent oflong nominal compounding in Tokyo Japanese. Studies in Language 11: 199-- 206. Twine, Nanette. 1988. Standardizing written Japanese: A factor in modernization. Monumenta Nipponic.a 43: 429--454. Uemw·a Yukio Cl:.fl'$:kt). 1997. On'in (Phonology). In Nilwnretto no gengo (flte Languages of the Japanese Archipelago), Kamel Takashl. (I,#'*"), Kono Rokuro (:i"PJ!lb'dl!l) & Chino Elich!. ('f-Jf:* ---) (eds), 211-235. Tokyo: Sanseldo. Ueno Kagetomi CUt ;f.tJ). 1980. Eigogoi no kenkyu (Research on English Vocabulary). 1bkyo: Kenkyusha UHABKS l1NESCO HigashiAjia BunkaKenkyii. Senta (.:;_;t;z::IJI'(J">'Tx1t"lf~-l!Y~-) (UNESCO East Asia Cultural Studies Centre). 1975. Shiryo oyatoi gaikokujin (Materials and Data on Early Meiji. Era Foreign Advisors). 1bkyo: Shogakukan Umegaki Minoru (~;i!.~) (ed). 1956. lngo jiten (Dictionary of Slang). Tokyo: 'Iokyodo. Umegaki Minoru. 1963. Nilwngairaigo no kenkyu (Japanese Loanwords: A Study). Tokyo: Kenkyusha Unger, J. MarshalL 1996. Literacy and Script Reform in Occupation Japan. Oxford: OUP. Unger, J. Marshall. 2004. Alternations of m and b in early Middle Japanese: The deeper significance of the sound-symbolic stratum. Japanese Language and Literature 38: 323-337. Ura, Hiroyuki. 1995. A metrical and optimality-theoretical approach to gemination and epentb.esis in Japanese loanword phonology. Osaka University Papers in IingliSh Lingui.~tlc.~ 2: 153-196.
Vance, Timothy. 1979. Non.~ense-word Experiments in Phonology and Their Application to Rendaku in Japanese. Ph.D. dissertation. University of Chicago. Vance, Tl.mothy. 1980. The psychological statu.~ of a constrai.rrt on Japanese consonant alteration. IJnguistics 18: 245-267.
225
226
Loanwords in Japanese Vance, Timothy. 1982. 0 n the origin of voicing alterati.on in Japanese consonants. ]oumal of the American Oriental Society 102: 333-341. Vance, Timothy. 1987. .An Introduction to Japanese Phonoiogy. .Albany NY: State University of New York Press. Vance, Timothy. 1996. Sequential voidng in Sino-Japanese. ]ournai ofthe As.~o.:iation of 'Teachers ofJapanese 30: 22-43. Vance, Timothy. 2005. Rendaku in inflected words. In \·bicing in Japanese, Jeroen van de \\'eijer, Kensuke Nanjo & Tetsuo Nishiliara (eds ), 89·.. 104. Berlin: Mouton de Gruyter. Vance, Timothy. 2007. Have we learned anything about Rendaku that Lyman didn't already know? In Current Is.~ues in the History atid Structure oj]apanese, Bjarke Frellesvig, Masayo .. shi Shibatani & John Charles Smith (eds), 153·.. 170. Tokyo: Kurosio. Vance, Timothy. 2008a. Ihe Sounds oflapatlese. Cambridge: CUP. Vance, Timothy. 2008b. Romanization and phonemicization of Japanese. In Japanese!Koreatl Linguistics, Vol. 13, Mutsuko Endo Hudson, Sun.. Ah Jun, Peter Sells, Patricia Clancy, Shoichi Iwasaki & Sung-Ock Sohn (eds), 81 ..·91. Stanford CA: CSLI. Vance, Timothy & Matsugu, Yuka. 2008. The !yf..fw/ asymmetry in Japanese loanwords. In Asymmetries in Phonology: An East-Asian Perspective, Haruo Kubozono (ed.), 139 .. ]<16. Tokyo: Kurosio. van Hout, Roeland & Muysken, Pieter. 1994. Modeling lexical borrowabllity. Language ·variation and Change 6: 39.. 62. Vasishth, Andrea. 1997. A model minority: The Chinese community in Japan In Japan's MitiOritie..~: Ihe Illusion ofHomogetleity, Michael Weiner (ed.), 108..·139. London: Routledge. Vendelin, Inga & Peperkamp, Sharon 2006. The influence of orthography on loanword adapta.. tions. Lingua 116: 996..·1 007. Vos, Frits. 1963. Dutch influences on the Japanese language. Lingua 12: 341·.. 388. Vovin, Alexande1: 1993. A Recotlstruction of Proto-Ainu. Leiden: Brill. \·Vatanabe, Shln'khiro & Hirato, Nobuo. 1985. 'Ihe relation bet\veen the perceptual boundary of voiceless plosives and their moraic counterparts and the duration of preceding vowels. Onsei Gengo 1: 1-8. Watanabe, Toshiro, Skrzypczak, Edmund & Snowden, Snowden (eds). 2003. Kenkyusha's Nev.· japanese-English Dictionary, 5th edn. Tokyo: Kenkyiisha. Weiner, Michael (ed). 1997a.]apan's},finorities: Ihe Illusion ofHomogenei~v. London: Routledge. Weiner, Michael. 1997b. The representation of absence and the absence of representation: Korean victims of the atomic bomb. In Japan's Minorities: The lllusi011 of Homogeneity, Michael Weiner (ed.), 79-107. London: Routledge. Weinreich, Uriel. [1953] 1963. Languages in Contact: Findings and Problems. The Hague: Mouton. Wells, J.C. 1982. .Accent~ ofBnglL~h. Cambridge: CUP. Wend<, Gunther. 1966. Ihe Phonemics of Japanese: Questions and .Attempts. Wiesbaden: Otto Harrassowitz. Winford, Donald. 2003. An Introduction to Contact Linguistics. Oxford: Blackwell. Yamada Yuichiro ( wffi ~ -l'!]l). 2001. Gaikoku eiga daimei no katakana hyoki ni tsuite: Hensen to sono shakalteki lmi (Foreign film titles and their Katakana transcription: Shifts and social meaning). Hiroshima ShUdai RonshU 41 (2): l-52. Yamada Yuichiro. 2005. Gairaigo no shakaigaku (The Sociology of Japanese Loanwords). Yokohama: Shunpiisha.
References Yamada Yli.ichiro & Nanba Kyoko (llt$tt:$-f-). 1999. Gaira:igo no hihyo: Saikin 50-ne.rikan no shinbunshiryo no kento (Criticism of loarrwords: An investigation based on newspaper content over the last fifty years). Hiroshima Shudai Ronshu 40(1): 143-181. Yamaguchi Masaya (ill t:l1j-\t). 2007. ShinbunkiJi ni okeru goi no jikanteki henkabunseki: goshu to no kankei o chiishin ni (Diachronic change in Japanese newspaper vocabulary with special reference to word types). In Hokoku 126: Kokyobaitcil no gaimigo- gcilraigo iikae teian o sa.~aeru dlc}sa kenkyu (Report #126: Loanword~ ln the Public Media- Basl.c Research on 'Paraphrasing Loanwords'), Kokuritsu Kokugo Kenkyiijo (f;!l:Y:f;Nlf.,~.~f) (National Institute for the Japanese Language) (ed.), 379-386. Tokyo: Kokuritsu Kokugo KenkyO.jo. Yamaguchi, Toshiko. 2007. Japanese Linguistics An Introduction. London: Continuum. Yamamoto Matsujiro (w*.t£.·*-1¥]3) (ed.). 1872. Shuchin haigo yakunO!Deutsch-japanisches Taschenworterbuch zum Gebrauche for Schuler, Kunstler, Reisende und Auswanderer. Nagasaki (Publisher unknown). Yamane, Noriko. 2001. sC dusters as complex segments: Evidence from the contrastive phonology of English and Japanese. In Issues in Japanese Phonology and Morphology, Jeroen van de Weijer &Tetsuo Nishihara (eds), 357-387. Berlin: Mouton de Gruyter. Yamane, Noriko & Shin'ichi, Tanaka. 2000. English Word-Final sC Clusters in Japanese Loanword Phonology. Phonological Studies 3: 37-44. Yamauchi Ikuo ( JJ FJ 1f }# ). 1972. Kanazukai no rekishi (The history of Kana usage). In Kliza nihongoshi 2: on'inshi, mojlshi (A Course in the History of the Japanese Language, VoL II: The History of Phonology and Orthography), Nakata Norio (If re-m.;;'<:) (ed.), 531--620. Tokyo: Taishttkan Yonekawa Akihiko (*Jl] o/l_,). 1997. Wakamono kotoba jiten (Dictionary of Youth Speech). Tokyo: Tokyodo. Yoshizawa No rio (t~ -~) & Ishiwata Toshio (;r.; Mlt*:.;tt). 1979. Gairaigo no gogen (Loanword Etymological Dictionary). Tokyo: Kadokawa Shoten
2.2.7
Donor word
Ainu aynu 67 iomante 66 .kotan 66 ruipe 66 susan1 66 tonakkay 66, 129 Arabic ld3iha:dl 43 lfatahl 4J 108 lbama:sl 87 lintifa:<,ial 43 lmrek:rel 9 lqur?aml 84
Beijing Chinese ap& R' lt 62 chaofan #f.i 64. 89, 129, 168
Ch6ngqlng :1:,~ 64 doubanjlang .a*£if 64, 84 dwmian ~w 64 Gao Xlngjian iii .ff(lt 65 gaoliang ilV 8, 64, 84 Hu Jiiii:iio 1!I'J i*.litf 65 jiiozi tsf:t 8, 62, 63, 64 l.amian :hi: iii 64. 168 maji3ng .$~ 64, 89, 90, 168 map6doufil .i+f.~.li.~ 6J 64
*
maquei~F-'l 64
mi.3nzi iii' t· 62, 64 pai .A* 64, s4 pill}in ;llt-·i 64 pu'ercha1t~if 64 Shimghai 1:.~ 64> 87 TAibei ~ ;:fr. 64
:*:;*
63, 64 will6ngcha .Jt. xiaol6ngbao ;J,~1!. 64.84 yapian~):f 62
yi- 1, 8, 64
zhaj&gm& flll:-l ii 64, 89,90
Cantonese cha1siu1 ~;Itt. 64 HeW1g1gong2 :i!):~ 64
gi.ps 39
lai~i 1 ~:.tt 64
gom 14. 38, 41, 42 Groningen 109 Gullit 88 haak 41 halogeen 40 ham 38, 41, 87 hop 41 jodium 168 kalk 39,107 kameleon 41 kamfer 39, 86, 90, u9 kan 40, 41, 168 kandelaar 35 karaat 39 katheter 39, U9 katholiek 33 kina 39,170 .kobah 40 .koffie 14, 35, 38, 40, 41, 86,
siu1maai 6 :!#1:* 64 wan4t:an 1 .t:-lf. 64. 84> 92, 97 yam2cha 4 i;z;f 64 C:r..ech Navratilova 84 Vlatava 85 Dutch aether see ether alcohol 39 alkali 39, 128, 140 Anderlecht 109 asbest 40 bier 28, 38, 41, 42, 140, 170, 171 blik 41> 107 boter 42 brandpW1t 149 Brussel122 cacao 41 calcium 40, 150 cement 39, 125 chemie 40, 42, 44> 8 7 cholera )8, 39 dans )8, 41 dek )8, 41, 107, 122 DenHaag 88 digitalis 39, 82 d.oek 38, 41, 8:1., 175 dok 38, 40, 41 dronken 42 Duits 38, 41, 68 elektriciteit 39, 132, 136 Enschede 87 ether 39,170 extract 28, 39, 108, 176 flanel 39, 41, 86 gas 28, 39, us, 138 Ghent 88
glas 28, 3.5. )8, 41, 42. 111, 157.168
149,168
kok 35, 40, 41 konuna 41 kompas 39 .kop 35 lamp 41 lens 28, 39 lepra 40, 42 letter 41, 122 lithiwn 40 lymfe 39 Maastricht 109 magnesium 40 marmot 38, 40 mast 38,41 mes 39 morfine 40, 86, 136 Nijmegen 88 nikkel 40, 122 ons 39 orgel )8, 41, 88
230
Donor word pek .p., 107 pest 39 piksee pek pistool 41, 153, 1il platina 40 pomp :l9, l.il pond .J5, 38, 39, 40, 41 pons 38,39 pudding 125 ransel 41 retort 39 reumatisch 142 roeper 41 saffraan 41 Schiphol 87 schop 40, 41, 87 serge 39, 41 siroop 41 soda 39 sop 42 spuit :19, 100 strychnine 40, 111, 112 thermometer 39 tinctuur 39 ton 39 trap 111 typhus 40, 82 Utrecht 98 van Gogh 88, 109 vernis 135 vet 41 zonda.g 42 English
-full, 141 -ism 1,141 -ship 141 4WD 189,190 accelerator 133, 134 accessibility 203 accessory 174 accountability 57, 200, 203,204 admission 117 after-sales service 149 agenda 203 agem 88 AIDS 88,189 alarm 150 album 176 Alice Springs 84 all 58; see also alright All Nippon Airways 189
Alps ;6, 142 alright .ss, 124 aluminium 176 ambition 117 America 140 1\merican 1.24, 149 American football 1-46 ammonia 90 amp132 analog(ue) 115 anarchist 56 anchor 104 and see copy and past.e, game
mul set, hit-and-run, Qt..A, shampoo and set and!or 1, 58, 104 android no animation 133 announce 137, 138 announcer 57, 144 •.1\ntwerp 68 AP 189 apartment 134, 150 apple 118, 154 applemint 119 appointment 132 apron 154 aqualung 107 Arizona l2J3 Armenia 68 Asi.an 150 asparagus 133 astr.ingeut 125 athletics 141 ATM 189 auction 106 Australia 140 authority 173 automatic 133 AV 188,189 axel 150 baby 141156 back office 201 backgammon 90 background music 149 ball 15;, see also baseball,
volleyball ballast 125 ballet 185 ballpoint pen 149 Band Aid 68 bandsman 142 Bangkok 68
bank 44 bar 128 barcode 155 barrier 157 base 146 baseball 55 basement 189 batter 104. 116, 174 batting 157 bead 110 bed 22, 51, 73.115 beefsteak 55, 124 beer garden 42 bell 69 bellboy 55 best 139 BigMac 133 bilingual 92 biotelemetry 153 birth control 146 biscuit 150 black ;8, 6o blanket 152 blizzard 150 blog 115 blouse .56 blue 58, 6o, 156 Blue Mountain 145 blues 99 BMW 12 board 125; see also .W!teboard body 145 Bombay 68 bonus 154 boom 140 boomerang 125 boot 110 boring 185 bowling 185 box 108,119 boxer 108 boxing 176 boy 55, 147, 156 brain trust 175 brake 107, 157 break 157 Britain 106; see also Great Britain brush 55, 109 bubble 82 bucket 55, no bu!fer 117 building 59, 132, 175
Donor word 2.31
bulldog 115 bus 11_5, 132 bushel 118 butter 42, 117 buzzer So, 129 bye-bye 58 ca.bbage 55 cabin 98 caddie 98 Cairns 104 Cairo 68 cake 107, 174 calendar 104 camera 44 9/. see also digital
camera camp 152, 178 can 40 candy 83 canoe 129 car 157; see also patrol car cardigan 135 career 98, 103 case 6o case by case 18.2 ca.~h 115 casual 148 catalog(ue) 97, 16S, 171 catch 1.56, 157 CD 189 cement 39, 125 centre/center 59, 6o, 61, 157; see also off-site centre chain 75 champagne 49 chance 156, 157 change 157 chatroom 88 check it out 125, 126 chef 72 chemistry 40, 42, 44 87 cherry 156 chess 181 chewing gum 42 chicken 119 child 144 Chile 172 chocolate So, 132; see also milk
chocolate Christmas 33, 145 Christmas Eve 149 Christmas tree 110 drde 141. daim 154
dass 59 classical 117 dear 103 dick 114 dub 171, 178 colla.borati.on 1.95 c..1lo(u)r 59, 187 Cohunbia 178 commercial message 190 commission 154 communique 174 company 145, 155 compartment 132 complex 145; see also mother
comple:c compliance 203 component 145 computer 104, 145, 182, 184; see also office computer,
personal computer, supercomputer concentr.ic phtg 133 concert 145 wncrete 145 condensed milk 141 condenser 145 condom So conductor 14.5 conference 97 connections 132 conscious 145 consensus 200 conso.rtiwn 194, 203, 204 consul 55 contest 145 continuity 148 contractor 145 control 138, 145; see also birth
control, remote control convenience store 130, 131, 133, 139,198,199 cookie 82 cooling-off 153 Copenhagen 68 copy So, 117, 142, 156 copy and paste 142 correspondence 136 cosmopolitan 130 cost 141 cotton 119 counselling 137 C\"'UIII 101 counter 14.5
counterpart 203, 204 couple 50 course 140, 157 Coward 104 crc!pe 50 crew 129 Croatia 67 cross cotmtry 144 cruise 107 Cubs 84 cuh 57 cunning 154 curry 100, 152 cut 97,140 cutlet no, 147 cuttir1g 174 cyde 138 darling 125 day 12, 156, 199, 205; see also
Valentine's Day decanter 83 decode 83 delicate 114 delivery 156 demo 132, 138 department store 133 depth 119 design .59, 187 dessert 50 deuce 177 diagram 136 dialt57 'liamond 136 didgeridoo 82 diet 57,114 differential gear 148 digest 181 digital 83 digital camera 144 dilemma 83, 175 disc jockey 57 Disney 181 dock 40 doctor 151 dodo 82 domain 185 door 104, 157, 176 double 138 down 58,141 dozen 124 drawers 110, 175 dredger 117 drive 175
232.
Donor word drug 140 Dublin 172 duet 9+- 181
duff 138 duffle coat 119 dummy 129 dutyfree 94 DV 189 DVD 82
ecology 132, 135 elevator 104, 176 empowerment 194 202, 203, 204206
enforcement 194 203,204 engagement ring 142 engine 211, 59, 144 157, 187 entertainment 136 equal 92 erotic 139, 141 EU 189
eucalyptus 133 eureka 184 eve 106; see also Christmas Eve event 200 Ever Ready Sharp Pencil 69 e>.'titing 1o8 exclamation mark 108 expert 108 extension 133 extra large 190 face 178 fadeout 72 family 145 fan 16, 56, 86 fashion 156 fastfood 57 fault 86 fax 12,58,59,108,188 Febreze 138 feminist 155 fence 86 fencing 181 festival 135 fever 172 file 14, 72 film 86, 107, 175 final 105 fmger bowl 185 Finland 68 first 58, 141; see also ladies first flash 107 FM 189
focus 72 fork 104 foundation 101 fraction 135 frank 132 free 149, 157, 205; see also dttty free, tax -free freelance 149, 150 friend 146, 156 front 148, 157 front desk 148 fruit 110 fry 144 fuse 86,177 fusion 181 gadget 97 gag 75, 115, 141 gallery 98 gamble 98 game 142, 157 g-.t.me and set 14'2 gangster 91 garden 105; see also beer garden gasoline 157 gate 185 Gates 185 gear 103; see also differential gear gelatin 88 gene.ra.l 145 general strike 88 Ge.neva. 68, 122 gentleman 88 Georgia 68 geranium 88 gesture 72, 175 get 55,199 getter 1.57 GI 57 gin 95
girl 147> 156
glass 42 global 205 globalization 154 glove 181 glucose 185 glycerine 124 go 155 goggles 82 gold 55 golden 156 golf .57
goods 110 Google 138 gorilla 149 Gothic 117 grandstand pla.y 149 grape 153 Great Britain 35, 53 green 58, 106, 149, 157 greenhouse 149 grey 129 grotesque 139 ground 101, 149, 157 group 6o, 106 guaranteed fee 98 guava 92 guide 156 guideline 57 guts 114, 157 hairpin 174
half 156 hammer 90 hamster 90 handkerchief 55, 124, 125 handle 154 happy 58, 116 happy ending 141 Harley-Davidson 100 harmonization 203 harmony 138 Harry Potter 146 hashish 109 hatchback 119 he 55 heading 177 healing 57 health 156 heart 141 hello 146, 157 Hepburn 124 high 140, 144 Hindu 83 hips 178 hire 103
hit 178 hit parade 146 hit-and-nm 142 hockey So, 100 hole 157 home 86, 135, 147, 198, 199 home helper 199 homepage 147, 154 honeymoon 44 hood 87
Donor word 233 horror 87 Hotchkiss Co. 68 hotdog 97 hotel 55, 59, 6o, 61, 129, 150, 156,187 hour 156 house 42, 55, 157 Hummer 90 hunger strike 91, 144 Huntingdon 91 ice cream 100, 149 idling 142 idol144 image 114 impression 106 in 1, 6o, 141; see also log-in incentive 57 inch 109, 178 incubation 202, 203, 204 incubator 201 inflation 133,135 informed consent 142, 205,206 infrastructure 133 initial118 innovation 195 install 136 instant 57 instructions 136 imtructor 136 instrumental 136 inter- 58 interchange 190 internet 194 interview 156, 177 intro 132 Iran 129 Iraq 6o iron 104 IT 189 Italian 148 J-pop 119 jack 107 jam 90 Jamaica 67 jamming 90 Japan Railway 190 Japanese 150; see also !-pop jazz 56, 115, 174 jeans 101, 144 145 jeep 57 jelly 88 jet 75
jet engine 175 jinx 154 jitterbug 125, 126 jogging 88 judge 114 juggle 118 juice 154 177 junction 106, 130 junkfood 57 kerosene 101 ketchup 117 key 157 kids 107 kilowatt 178 kitchen 75, 106, 119 Klaxon 68 know--how 57 lace 100 ladies first 141 lady 190 lambswool 142 land 156 lane 101 layout 182 league 56; see also Pacifu
League leather 87 Leeds 88 leggings 125 leisure 57 lemon 89 lemon squash 147, 176 lemonade 124 lesson 106, 119 level 141 lever 157 lifestyle 57 light 157 lighter 176 lincoln 172 line 100; see also mainline lingerie 50 liquor 117 lisbon 68
listu literacy 203 live 157 Llanelli .'12 loan 101, 154 location 156 lock 91 log-in 195 Lolita 145
London 67, 74 lonely 101 long .140 look 156, 157 looks 97 love 115, 156 lover 144 lucky 58, 142 lucky seventh 142 lute 100 Mac 133, 140; see also Big Mac MacDonalds 133 Madame Tussauds 142 magazine 85 magenta 88 magic 149 magnetic 141 mail address 147 main 185 Maine 185 mainline 75 maintenance 135 make-up 149 maker 6o mama ;8, 90 man 69, 75, 142, 157; see also bandsman, salaried man, spokesman manifesto 154 manner(s) 157 manpower 143 map 97 marathon 57, 87 mark 14 104; see also
exclamation mark, question mark mashed potatoes 142 mass commmtication 148 match 171 Matthews 94 McShake 133 measuring spoon 141 medley 100 memo 138 memory stick 148 merit 200 message 117; see also
commercial message metabolic syndrome 133 metronome 56 mic 57, 132, 156 micro- 140 microphone 57
234 Donor word middle 1.57 mike 132 milk chocolate 151 milkshake 87, 175 mince 55 mini- 140 Mini
Mountain muffin 119 multi-· 140 Mwnbai 68 my 1,58,60,140,198,199 naive 154 name 178 native speaker 148 natural117 needs 57, 89, 110, 154 200 NEE'f 189 neglect 139 negligee 50 net 132; see also
internet net radio 146 network 104 neutrino 140
new 156,178 New York 6o, 172 ne\\--s 57, So, 100 newscaster 149 NG 189 NGO 189 nickel cadmium 1-46 nighter 8 no 140 no comment 58 non- 58,140 nonsense 58 normalization 154, 205 notebook 149 nothing 117 nougat 50 now 139 number 6o nut 110 nylon 150 oar 102 OB 140,189 ocean 190 odds 114 of 12, 13, 6o off 58, 115, 128, 14J, 202, 204; see also cooling off off the record 1<12 off-site centre 202, 203. 204 office 190; see also back office office computer 148 offside 97 OK 58 old debt 55 oldmiss 8 omelette 50, 110 on the air 142 on the rocks 141 one 1, .58 open 1o6 option 106 original 6o out 55, 57, 58, 156, 157; see also
check it out outsourcing 142, 201 oven So over 1 over- 140 overcoat 149, 173 overture 103 Oxbridge 119 Pa.citic League 147 Page 185
page .59, 109, 168, 178, 18.5, 187, 188; see also homepage paint 175
pair 157 panic 138 panther 89 pant~ 107, 144 papa 58 paper see sandpaper, toilet
paper, tracing paper par 104> 134 parody 139 part time 149 palty 72, 156 patrol car 144 pedantic 141 pen 149,157 penalty kick 190 pending 58, 153 perm(anent) 57, 133,134 personal computer 146 persuade 55 pet 156 phone 157 phrdse 106 piano 174 pierce 104 pingpong 125 pink 91, 139, 156 pipe 155 pitcher 117 plastic model 147 platform 86, 135, 175 play 55, 156; see also
grandstand play plus 157 poached egg 142 pocket 144 Poe 83 point 59, 6o, 61, 149, 157 Poland 182 police 105 policy 130 pool 107 pop 143; see also !-pop pork 104 pose 157 post 56 potato 11, 12, 144; see also
mashed potatoes pouch 100 powder puff 149 PR 189
Donor word 2.35 precast 145 present .59 presentation 134 price 141; see also road pricing,
Tokyo Stock Pr'.ce l1ltlex pride 91 printer 1.04 privacy 57 producer 177 professional 6o program(me) 6o proletariat 6o propaganda 6o propeller 135 propose 137 prostitute 6o PTA 189 pub 106 public affairs 104 public involvement 203 pudding u5, 126, 129 ptuna 93 purple 82 pyjamas 141 Q&A 189 Qantas 92 quark 92 quarterly 92, 176 question mark 181 quintet 92 race 59 radio 8, So, 145, 146, 173 rally 129 range 101 Rangoon 68 ranking 91 rape 185 reasonable 58 recipient 202, 203. 204 record 57; see also
off the record, tape recorder recreation 194 recruit 57 recycle 194 recycling 138 reggae 185 register 132 rehabilitation 133, 199 remote control 144 renewal168 resort 144 rest 150
restaurant 50, 126 restructuring 133 review 148 revue 181 rhythm 107 ribbon 129 rice 48,1.54 risk 200 road 157 road pricing 203. 204 roast 124 roots 57, 110 rule 91 sacdtarin 116 safe 55 salad 6o, 112 salad bowl 182 salaried man 141 sales 84; see also lffter-sales
service sand paper 149 sandwich 133. 147, 176 sapphire 10:1 sash 109 savannah 89 save 84 scab 114 scan 157 scanner 89 scene 85, 99 sd1eme 200 science 154 scoop 40 Scotch Tape 97 scrambled egg 141 screen 106 screwdriver 149 Sears 104 season 84> 141 second hand 144 section 140 self-service 149 semi- 140 sense 95 series 59 service 6o, 155, 156, 199, 205; see also
lffter-sales service, self-service set 28, 58, 59, 187; see also
game and set, shampoo and set settlement no
severe 139 sewing madline 149 sex 108, 146, 156 sexual harassment 144 Sel."}' 57, 108 shade 87, 175 shaker 1.81 Shakespeare 172 shampoo 87 shampoo and set 142 shareware 103 shear 103 sheet 110 shepherd 87 shirt 55. 110, 116' 124 shoe 143 show 100, 156, 178; see also
strip show showboat 178 shower 103 shutter 156, 174 side 157; see also SUfJfJiy SUie sign 137. 156, 157 Silicon Valley 100 silver 140, 1.55 simulation 205 singles 142 size 28, 59, 157, 187 skateboard 144 sketch 114 ski(ing) 60, 129, 141, 178 skin 141 slim 140 slipper 176 slogan 56 slugger 117 small 157 smart 15.5 Smith 115 smog 114, 149 smoked salmon 141 smoothie 87 snob 114 soap 156 soccer 57, 96, 97 soft 149 189 soulful 185 soundtrack 145 soup 42 southpaw 185 spa 99
sos
236
Donor word Spain 68 speaker 148; see also nam-e
speaker spelling 141 spike 157 spin 1.29 spirit(s) no spokesman 108 sponge 34> So sponsor 57 spoon 101; see also measuring
spoon sport(s) 59, 6o, 61, 187; see also Ministry of Educatiott,
Culhtre, Sports, Science and Technology sportswear 103 sprltzer 117 square 176 squash 92; see also lemon
squash stadium 176 staff n, 115, 194 stainless steel 101, 149 stal1144 stand 157 staple fibre 148 star 178, 184 Starbucks 138 stardust 148 start 6o status 107 steak 107 steward 104 stewardess 177 Stewart 104 stick 83, 107, 156, 171; see also
memory stick sticker 83 stop ;9, 145, 151, 155, 156 stopwatch 176 stove 155 straight 110 straw 99 stress 57, 194 strike 57, 107, 133; see also
general strike, hunger strike, three strikes strip show 57 struggle 118 Stuart 104 studio So studless tyre 149 style .59
Style Council 14.5 stylish 115 success 156 suit 56, uo suitcase 110 summer 145 sun 190 sunglasses 57, 141 super 140 supercomputer 146 supermarket 59, 148, 187 supper 55 supplement 133 supply side 203, 204 surfmg 125 surveillance 203 sweater 92 Sweden 67, 181 swimming 90; see also
synchronized nvimming sw.itdl 92 switcher 75 symbol 90 symposium 105, 134, 18.2 synchronized swimming 142 system 59, 6o, 61,145,187, 201; see also 1bkyo
Broadcasting System tackle 118 taco 143 talent 149 tanki.ni 149 tape recorder 147 tartan 82 task force 194> 203,204 tattoo 72, 82 tax.. free 119 ta.'ti 107, 108, 182 tea 154> 175 team 59, 6o, 83, 175 teen 140 tel(ephone) 12, 28, ;8, 59, 187,188 tele- 157 television 59, 6o, 61, 133, 149,186 ten 157 terrorism 6o, 61, 134 Texas 108 teJi.t 107, 108, 176 thankyou 58 the 1, 12, 13, 58, 59, 6o, 75, 140, 141,188 theatre/theater 87
thinner 89 thoroughbred 21, 114, 115 three 58 three strikes 141 thriller 1o6 throttle nS ticket 117 tie 6o time 156; see also part time tip 114 tire see tyre TOEIC 189 toggle 118 toilet paper 100 Tokyo Broadcasting Sy~tem 190 Tokyo Stock Price Index 190 tongue 125 top 6o, 146 toss 157 tour 103, 145 town 22. traceability 201, 203, 2.04 tracing paper 147 transparency 1:16 tree 175; see also Christmas
tree triangle 135 trouble 138 truck 112 trump 154 tuba 181 tulip 178 tweed 111 tweeter 111 twelve 111 twenty 111 twin 111 twirler 110 Twitter 111 two 58 two-piece 175 type 28, 58, 59, 187 tyre 157, 174; see also shtdless
tyre UllO 188, 189 ultra 140 unbelievable 58 UNESCO 189 unique 89 up 58, 6o, 137, 141, 146; see also
make-up upgrade 119 utopia 150
Donor word 237 vacation 50 Valentines Day 142 variable 145 variety 144 velvet 35 velveteen 1J5 veneer 174 veteran 85 vibrator 150 video 153 VIking 69 violin 85, 183 virgin 157
vision 200 vitamin 175 vocabulary 175 volley 97 volleyball 97, 100 volunteer 181, 194 vomiting 172 VTR 189 wafer 143 waffle 118 waist 178 waiter 92 Wales 92 walk 157 wash 69 Washington 6o, 91 waste 125 water 156 water polo 92 wave 176
we
56
weakness 55 web 75 website 195 weekend 92, 154 Western 157 Westminster 130 whale watching 93 whisk(e)y 172. 176 whistle 93, 118 white 12, 93. 124, 156 Wichita 92 wide 156 Wikipedia 79, 132, 138 wine 92
Winnipeg 75 wire 103
Wisconsin 97 Women'sli.b 142 Worc.ester sauce 92 word processor 144
work 157 WTO 189 Xerox 139 yacht 150 Yale 76 Yangon 68 }"Ud 93, 125
yeast 93 Yeats 181 yell 154 yellow card 93 yes 58 you 1, 6o, 99, 134; see also thank you zero So zero emissions 201 204 Farsi lrehmredi:n~:>:dl 107
IGoml 84 lteho:lll 172
Frendt
aIa carte Aix
48, 49
122
Antibes 122 Asterix 122 atelier 49 avec 50 baguette 121 ballet 49, 185 Balzac 121 Bariquand & Marre 68 beige 107 bifteck 49> 124 Bobigny 91 Boulogne 91 Bretagne 10.5 Caen 90 Camus 98 Cannes 89, 90 Caran d' Ache 121 centimetre 28, 49, 133 Cezanne 89 champagne 49, 50 Chanel 141 chapeau 49, 97, 122 Chartreuse 99 chic 49 cognac 49, 50 concours 48, 49 consomme 49 coup d'etat 49 cr~pe 50, 153
crl!pe de chine 149 croissant 49, m croquette 112, uS, 125 cuisine 89, 94 culotte 97 deja. v11 s5 dessert 50 dessi.t1 49, 101, 122 Dumas 99 Eiffel 122 en croftte 121 encore 49, 107 enquete 49, 121 escabeche 121 esthetique 121, 133 etiquette 48, 49,50 fiance(e) 173 filet 49,95 foie gras 111 France 140 Geneve 68, u2 genre 49,87 gramme 28, 49, U9 grandp.rix 49 gratin 48, 49 Guadeloupe 121 hors-d~uvre 49, 98
Jules 99 j upon 49, 140 kilogramme 49,95 kilometre 28, 49 la 1; see also ala carte lingerie 50 litre 49, 122 Louis-Philippe 121 luge 99 marron 101 Marseille 107 Matisse 122 mayonnaise 50, 89, 133. 141 mecenat 50, 194 metre 28, 49. 61, 122 Metz 122 millefeuille 50, 98 millimetre 28, 49, 132 moin! 111 neglige 50 nougat 50 omelette 48, 49, 50, no onomatopee 168 Paris 67, 87, 178 Pau 83 Piaf 122 pi.ment .50, 129, 155
238
Donor word Poitiers 111 pornographic 102 potage 109 profile 148 Quebec 121 ratatouille 87 reportage 49, 105, 122 restaurant 50, 126 roman 171 sabotage 49, 138 salon 156 Saussure 99 Seine 89 sommelier 50 Suisse 140 toile no Toulouse 83 trompe-r oeil 98 Tuileries 94 un-deux-trois 98 vacances 50, 101 Verdun 101 Versailles 181 vinyle 48, 49, 85, 150 Gaelic
loch 109 German
Adenauer 105 Allergi.e 28, 52 Arbeit 28, 52, 135 Arbeiter 150 Asparagin 52 Bach 87,108 Beethoven 183 Berlin 105 Bett 51 Bismarck 172 Bombe 53 Bruch 109 Buchner 109 Buchner 109 Chloroform 52 Dachshund 53 decken 52 Demagogic 52, 53 Dohran 69 Doppelganger 91 Dortmund 105 Diisseldorf 99 Eisbahn 53 Eisen 52, .53 Eisenach 108
Energie .52, 140, 1.50 Felix 122 Gaze 51, 52, 99, 129 Gel 53 Geliinde 52, 53> 105 Gewalt 53 Goethe 99, 172 Hamburg 105 Haschisch 109 Hindu 83 Hoden 51 Hormon 53,107 Hiitte 52, 122 Hysteric 99 Ideologic 52, 53 Impotenz 52, 133 Jena 93 Jodel 93 Joghurt 102 Kapsel 52, 97 Karies 5:1 Karte 52, 53 Kartel 52,102 Koks 52 Koln 98 Kot 51 Leipzig 72 Liebe .52 Liechtenstein 109 Lwnpen 53 Lymphe 40, 51 Mad1 122 Miirche.n 52, 87, 141 Marx 140 Messe 53 Miinchen 67 Neurose 52, .53, 99, 100 Niirnberg 98 Oktan 144 Onanie 53.156 Parenthese 134 Ffalz 89 Ffennig 89 pH 188,189 Pornographic 102 Potenz H2 189 ROntgen 51, 52, 98 Rucksack 52, 129 Ruhr 91 Saarland 105 Schanze 52, 53, 72, 181 Schlaf.~ack 52, 53
Schnapps 122 Schnorchel .53
schon .52 Schutz 99 Schwarzkopf 107 Seil 52, 53> 84 Seminar 52,1.32 Spur 168 Stock 28 Si.iskind 99 Synthese 99 Thema 28, 52, 99,194 These 53.105 Tripper 51 Tuberkulin 52, 82 Ttibingen 99 Typhus 40, 51 Urethan 28, 53> 97 Vakzin 52, 88 Virus 53 Wagner 105 Wedeln 85 Weif~l">ier 85
Wien 85 Wolfsburg 85 Wuppert.al 85 Wiirttemberg 177 Zeichen .52 Zeppelin 88 Ziirich 99, 109 Greek ll 75
Hawaiian
Oahu 84 Hebrew lxanuka I 87 Hungarian Gyor 99 Nagy 84,109 Italian
bolognese 91 cello 43,88 d.a Vinci 85 espresso 43, 94 focaccia 88 forte 43 Gigli 91 lasagna 91 macaroni 157 Marco Polo 183 Napoli 67
Donor word 239
scher1.o 72
fides 32 manna 32,33 oratio 32 pater 169 spiritus 32
spaghetti 43, 157 tiramisu 82 Venezia 85
Tromse~ 98
Korean
Polish
chijim ;:<] ~ 67, 89 ch'ima :'<.]P} 67 dr'imajogori ::<],}jJ.:;Ii!.] 67 chogori ~3li!.l 67 dr'onggak ~Zj- 67 chuch' e "f-;>JJ 67, 89 han'gQl ~~ 67, 84 Inch' on 'Zl ~ 66 kalbl t);lll 67, 84 Kim ChOng- il {i ~ ~ 65 Kim Ok-kyun -il4it 66 kimch'i ·tl :71 67, 84> 12.9 kisaeng 7]·\!1 67 onmun 'tl 67 P'yongyang ~ oJ: 66, 84, 90 paji il}.:>:] 67 Pak \If 12.3 Pan Ki··mun l~)-7]~: 65 pap 'F.f 12.3 pibimpap ll]llJ W 67 ppongtchak .J!!L·!lj- 67, 84 pulgogi ~317] 67,84 Pusan ¥ ~ 66 Sejong All* 66 ssampap -¥.1-W 123 Taegu cJIT 84 t'aeg(lkki '6il"t71 97 t'aekwondo blli!£ 67, 84 tchigae ~ 7] 67, 89 123 tolsot' ttok 1;Jij 123 wOil ~ 67,92 YiMyong-bak 0 l~llf 65 yukhoe ~ ~ 67, 93, 98
Cz~stochowa 101 Gdansk 109 Lech 101, 109 t6dz 92,109 Wal~sa 92, 109 Warszawa 67
opera 43
pasta 43· 129 pizza 43 salami 43
*
¥*
I..atin apriori 43 anima 32 ave maria 32 candela 34> 35 catechismus 32 deus 32 ecdesia 32, n1, 112 ego 43
Norwegian
Portuguese alamblque 124 altar 32 8.1~0 :12 banco 31 baptismo 32 bispo 32 botii.o 28,34 calcador 34> 152 capa 33, 34, 35, 123, 151, 152, 168,173 capitao 34 carta 34> 35, 151, 152, 168 Castela 33, 34 catholica 170 confC.ito 34 confissao 32 copo 34> 123 credo 32, 111 crlstao 31, 32, 111, 129, 170 Cristo 32, 33 cruz 32 espada 34 evangelho 32 filho 169 frasco 33> 34 gentio 32 gibao 28,34 Grecia 68 Rolanda 34> 3.5, 38, 68 h6stia 32 inferno 32 ingl~s 33> 34, 68 irmao 32, 3117o jarro 34 marmelo 34 martir 32, 170
meias 35 mirra 34129 mlssa 32, 33, 169 natal 32, 33 padre 32, 111 pao 28, 31, :lz, 65 pascoa. 32 Portugal 34, 35 profeta m, 169 raxa 33,34 Rond6nia 101 rosario 32, 33 sabao 35 sacramento 32, 111, 112 sacriffcio 169 santo 31,32 Sao Paolo 101 senhor 34> 35 tabaco 28, 31 34 35, 140, 168, 170, 173> 178 temperar I tempe.ro :z8, 33> 34, 35,168 tinto 3:1,34 vaca 34> 35 veludo 34> 35, 170 video 34> 35 Russian ApxaHrerrbcK 109 AxMaTOBa 109 6ana.11a~:Ka 45, 46, 129 Ee11apyc:& 93 fl011biUfBR KH 46, 85 6opll\ 46, 47, 107 Epe)KHeB 87, 109 BJia~OCTOK 85 BO,Il;Ka 28, 46, 47, 85 rJiaCHOCTI> 46,47 lpoMI>IKO 98 fpysiDI 68 ,ll;OMOt:t 46,47 EKaTep~m6ypr 93 Elrb~p;m 88 MBaH 85 JIKpa 28, 46, 47, 154 HHTelil!HreH~ 28, 46, 47, 133,134 KasaxcTaH 68, 109 KaMIIaHID.I 46, 47, 133, 134 KaM'laTKa uo
KHpfRS 68 Kosp;mqes 94 KolThiMll 98 KOM6JillaT 4.6
240 Donor word 98 46,93
KoMm!TepH 46
CMhJCJIOB
K}'llaK 122 JleHHH 172
COBeT
CoJiiKCHHJ:V>IH 181
C6rdoba 85 el Niiio 91 fiesta 72
MeWJe~eB 94
CII}'THHK 46
Guatemala 92
MocKBa 68, 85 HosocKuHpcK 85 HOpMa 28, 46, 47,
Tatlra 46 TO'IKa 46 Tpoji:Ka. 46 T}'HApa 46
hiniesta 34. us jabon 34,35
0XOTCK
91
87
rrepecTpotl:Ka 28, 46, 47 IIepMI> 109
IIC'IKa 46 III1)JOiKKH 46
~apb
Ilynm 94
IIIocTaKOBH'I 109
fi}'lllKUH 93, 109 PocTpoiiOBH'I 123 pyccKml: 46
Serbian
PJisaHb 93 CaMOBap 26,129 CKBY'I 45> 46
46, 47, 72,109
Juan 87
Llosa 91 medias 34. 35 Pablo 106 paella 91 Paraguay 181 Sevilla 68, 81, 85 Villa 81 xabon 35
Mm~omeBHh 123
Spanish candela 34
Swedish Helsingborg 91 Norrkoping 99
Loanword index
'This index is in two parts. Part llists loanwords in the Roman transcription used throughout this volume; Part II lists the same loanwords in Japanese script.
A aa 140,141 aateru 39, 170 abemaria 32 abeQku 50 adenauaa 105 adorniQSoN 117 adoresu 147 afeaazu 104 afumaatova 109 afumadiinejaado 107 afutaasaabisu 149 afUiaaseerususaabisu 149 aheN 62 ai 6o aiaN 104 aidorlNgu 142 aidoru 144 ainu 67 aisii 190 aisubaaN 53 aisukw·iimu 100, 149 aisukuriN 100 aitii 189 aizeN _52, .53 aizenaba 108 ajeNda 203, 204 ajia.N 150 akam.'tabiritii 57, zoo, 203, 204 akuaraNgu 107 akuseQsarii 174 akllSeru 133, 134 150 akuserureetaa 133, 134 akllSesarii 174 akuseSibiritii 202, 203, 204 amefuto 146 amerika 140 arnerikaN 124,146,149 arnerikaNfuQtobooru 146 amerikaNkoohii 149 ana 189
anaakisuto 56 anadoru 144 anarogu 115 anaUNsaa 57, 144 anatiNsaaaidoru 144 anaUNsu 137, 138 aNbariibaburu 58 aNbiSoN 117 aNderurehito 109 aNdooa 1, 58, 104 aNdoroido 110 aNdutorowa 98 aNguru 135 aninla 32 aninle 133 aninleesoN 133 aNjo 32 aNkaa 104 aNkeeto 49, 121 aNkoon1 49, 107 aNmonia 90 aNpt:L 132 aNtiibu 122 aJII1:owaapu 68 apaato 134, 150 apaa.tomeNto 1.34 apoiNto 132 apuriori 43 aQpu 58, 6o, 137, 141, 146, 149 a.Qpugureedo 119 aQpuru 118, 154 aQpurumiNto 119 aQraa 73 araamu 150 arakaruto 48, 49 arerugii 28, 52 arisusupuriNgusu 84 arizona 128 arubaitaa 150 arubaito 28, 52, 135 arubamu 176
aruhaNgerisuku 109 arukari 39,128,140 arukooru 39 arumenia 68 artmtiniumu 176 aruminyuumu 176 arupusu 56, 142 arutaru 32 asubesuto 40 asupara 133 asuparagasu 133 asuparagiN 52 asureciQku 141 asuterikllStt 122 asutoriNzeN 125 asutorlNzeNto 125 atorie 49 auto 55, 57, 58, 156, 15 7 autokoosu 157 autoseQkusu 156 autosoosiNgu 142, 201 awaa 156
B baa 128 baaji.N 157 baajiNroodo 157 baa.k:oodo 155 baasukoNtorooru 146 baburu 82 bageQto 121 baibai 58 baiburaamu 150 baibureetaa 150 baibureetaaaraamu 150 baikiNgu 69 baioriN 85, 18 3 baioteremetorii 153 bairiNgaru 92 baito 2.8, 52, 135 bakaNsu 50, 101
242 Loanword index bakecu .S5, 110, 171 bakeesoN .so bakeQto 110 baNdo 142 baNdoeedo 68 baNdornaN 1.42 baNgoQho 88, 1.09 baNko 31 baNkoku 68 baNku 44 bapuCizumo 32, 33 baputizumo 32 baQfaa 117 baQha 87,108 baQku 1.49, 201 baQkugurauNdomyuujlQku 149 baQkugyamoN 90 baQkumyuujiQku 149 baQkuofisu 201 ba.Qtaa 104, u6, 174 baQtiNgu 157 barad(ml 144 baraeti.i 144 baraetiiaidoru 144 bararailta 45,46,129 barasu 125 barasuto 125 baree 'ballet' 185 baree 'volleyball' 97, 185 bareebooru 97, 100 bareJI.'taiN 142 bareNtaiNdee 142 baria 157, 205 bariaburukoNdeNsaa 145 bariafurii 157, 205 barikaN 68 barikoN 14.5 baruzaQk.u 121 basu 115, 132 basukoN 146 bataa 42, 117 bateNreN 31 bateree 31 batereN 31, 32, 33, 111, 170 bateru 31 bea 146 bebii 156 bebiisaakuru 143 bebiisutoQpu 156 beeju 107 beesu 146 beesuaQpu 146
beesubooru 5.5 beetoobeN 183 beetooveN 183 bejitaburu 142 beneCia 85 beniya 174 beQCiN 135 beQdo 51, 73, 115 beQdotaUN 22 beQto 51, 115 beraruu8i 93 berubeQCiN 135 berubeQto 35 berudaN 101 beruriN 105 besuto 139 beteraN 85 blagaadeN 42 biblNba 67 bideo 153 bifuteki 49, 1.:24 b.ii 12, 13, 86 biidoro J4, :15 b.iiemudaburyuu 12 biifusuteeki 1.24 biiru 28, 38, 41, 42, 140, 170, 171 bii~'li.N 189, 190 biizu 110 bija 81 bijoN 200 bini 130 biniiru 48, 49, 85, 150 b.iniirunairoN 150 biniroN 150 biniru 130, 139 biQgU 133 biQgwnaQku 133 biroodo 34_, 35, 170 biru 58, 59, 132 birudiNgu 59, 132, 175 bisukeQto 150 bisumaruku 172 bisupo 32 bisuteki 55, 123, 124 bitamiN 175 biya 42 biyagaadeN 42 bobinl 91 bodii 145 bodiikoN§asu 145 bodikoN 145 bokiSiNgu 176 bokusaa 108
bokuSiNgu 176 boNbe .53 boNbee 68 boodo 125,144 booi 55, 147, 156 boonasu 154 booriNgu 185 booru 'ball' 97, 125, 149, 157 booru 'bowi' 182,185 boorupeN 149 boorupoiNtopeN 149 bootere 31 bootoru 42 boQkusu 108,119 boraNtia 181, 194 boree 97 boroneeze 91 borufusuburuku 85 boru8ebiki 46, 47, 85 boru8iti 46, 47,107 botaN .28, J4, :15 bui 86 buitiiaaru 189 buQku 149 buQpaat.aaru 85 buQseru 118 buraN·dopiNto 149 buraNkeQto 152 buraNko 35 buraQku 58, 6o buraQsu 109 bura8i 55, 109 burausu 56 bureeki 107, 157 bureeN 175 bureeNtorasuto 175 burejinefu 87, 109 buriki 41, 107 buriQki 41, 107 buriseru 150 buriteN 106 burizaado 150 burizaadoakuseru 150 burogu 115 buroQgu 115 burudoQgu 115 burudoQku 115 burwnaN 145 buruQfu 109 burutaanyu 105 burutawa 85 burtnl 58, 6o, 145, 156 buruudee 156
Loanword lndex 2.43 buruumatiNteN 14.5 buruusu 99 buryuQSeru lZ2 buucu no buumeraN 125 bmuneraNgu. 125 buumu 140 buuroonyu 91 buzaa So, 129 byuuhinaa 109
c caa 46, 47, 72 caari 46, 47, 109 caaru 46,47 ceQperiN 88
cuaa 103, 145 cuaakoNdakutaa 145 cuakoN 145 cuberukuriN 52, 82 ClJ.eNtii 111 cuerubu 111 cuiido 111 cuiitaa 111 cuiN 111 cuiQtaa 111 cUNdora 46 curii no, 175 cuu 58,175 cuupiisu 175 cuuruuzu 83
c caahaN 64, 89, 129, 168 Ca.a.Suu 64 Ca.idoru 144 Ca.irudo 144 Ca.irudoaidoru 144 caNpoN 63 CaNSU 156, 157 caNsubooru 157 caQto 88 caQtoruumu 88 ceeN 75 eekira 125, 126 eeko 72 CeNji 157 ceNjirebaa 157 ceNsutohova 101 eerii 156 eeriibooi 156 eero 43,88 cesu 181
cifusu 40, 51, 82 Cige 67,89 Ciimu 59, 6o, 83, 175 cijimi 67, 89 CikeQto 117 cikiN 119 Cima 67 cimaeogori 67 CiNki 39 ciNkicuuru 39 CiNta 3134 CiQpu 114 Ciramisu 82 Ciri 172 eoe caNhoa 65 cogori 67 eoko 132 eokoreeto So, 132, 151 CoNON 64 CoNgaa 67 cubiNgeN 99 cu.Ce 67,89 cui.rurii 94 ctmiNgamu 42 cuuriQhi 99, 109 cuuriQpu 178
D daariN 125 daariNgu 125 da.asu 124 dabiNCi 85 daburiN 172 daburu 138 daburyuu 12 daburyuu8ii 56 daburyuutiioo 189 dafu 138 dafuru 138 daieQto 57, 114 daijesuto 181 daiya 136 daiyaguramu 136 daiyamoNdo 136 daiyaru 157 damii 129 damoi 46,47 daNSU 38,41 daQfurukooto 119 daQkusufuNto 53 dauN 58,141 dee 12, 142, 156, 199, 205 deesaabisu 199, 205
deeta 185 defa.reNSa.ru 148 defareNSa.rugia 148 defugia 148 dejabyu 85 deja.vyu 85 dejikame 144 dejitaru 83, 144 dejitarukamera 144 dekaNtaa 83 dekoodo 83 dema 52,53 demo 132, 138 demoru 138 depaato 133, 134 depaatomeNt:o 133 depaatomeNtosutoa 133 deQki 38, 41, 107, 122. deQpusugeeji 119 deQsaN 48, 49, 101, 122. deribarii 156 deribar.iiherusu 156 derikeeto 114 deseen1 50 desuku 148 deusu 32 dezaato 50 dezaiN 59, 187 diibui 189 diibuidii 82 dijeriduu 8.2 dijerizuu 82 dijitaru 83 direNma 83, 175 disukujoQkii 57 dizunii 181 doa 104 157, 176 doaeNjiN 157 doicu 38, 41, 68 dokutaa 151 dokutaasutoQpu 151 domeeN 185 doNtaku 42 doodoo 82 doonacu 147 dooraN 69 doQku 38, 40, 41 doQpe.rugeNgaa. 91 doraibaa 149, 157 doraibu 175 doraQgu 140 doreQja.a 117 doroNkeN 42
244 Loanword index doru 38,144 dorutomu..'lto 10.5 dyueQto 94 181 dyurna 99 dyuQsenl.dorufu 99 dyuusu 177 dyuutiifurii. 94 E
ea 142 ebaaredi 69 ebaaredisaapupeNsiru 69 eCi.keQto 48, 49, so eebui 189 eeenuee 189 eejeNto 88 eepii 189 eeru 154 eeteru 39 eetiiemu 189 eezu 88,189 efuemu 189 ego 43 ekateriNburuku 93 ekeresia J2, 111, 112 ekisaitiNgu 108 ekisu 28,39,108,176 ekisukurameesoNmaaku 108 ekisupaato 108 ekisutorakuto 39 eko 72, 132, 135 ekorojii 132, 135 ekusu 122 ekusukurameesoNmaaku 108 ekusute 133 ekusuteNsoN 133 emiQSON 203. 204 emu 12, 13, 55 eNdo 141, 142 eNdoraN 142 enerugii 52, 140, 150 enerugiiyuutopia 150 eNetopia 150 eNfoosurneNto 194, 203,204 eNgeeji 142 eNgeejiriNgu 142 eniSida 34 128 enisuda 34 eNjiN 28, 59, 144, 157, 187 eNjiNsutooru 144 eNkuruuto 121 eNpawaame!;1:o 194, .202, 203, 204,206
eNstilieede 87 eNsuto 144 e:~~;'taateeNmeNto
136
eNtame 136 enu 89 enuiisii 190 em~ii 189
enujiioo 189 enutiitii 190 epidora 12 epuroN 154 eQCi 191 eQferu 122 eQgu 141, 142 eQkusu 14 eQkusueru 190 erebeetaa 104, 176 ereki 39, 136 erekiteriSiteeto 39 erekiteru 39, 132, 136 erekiteruseeriteeto 136 ericiN 88 eroCiQku 1:19,141 eroi 139 eruniinyo 91 esukabeesu 121 esuooesu 189 esupureQSo 43, 94 esute 133 esutetiQku 121, 133 ev\'aNzer.iyo 32 F
faasuto 58, 141 faburiizu 138 faburu 138 faibaa 148 fainaru 10.5 fairu 72 famikoN 145 famirii 145 famiriikoNpyuutaa 145 faN 16, 56, 86 faNde 191 faNdeesoN 101, 191 faQkusu 12, 58, 59, 108, 188 faQSON 156 faQsoNherusu 156 farucu 89 fasutofuudo 57 fataha 43, 108 fauN
feminisuto 15_5, 175 feNslNgu 181 feNSU 86 feriQkusu 122 festl 135 fesutibaru 135 fiaNse 17.3 fiesuta 72 fiNgaabooru 185 fiNraNdo 68 firumu 86, 107, 175
fivuru 172 fokaQea 88 fookasu 72 fooku 104 fooruto 86 forute 43 fowagura 111 fujlNtao 65 fuaN 'fan' 86 fuaN juan' 87 fueNSU 86 fuirurnu 86, 175 fuooruto 86 fuQtobooru 146 furai 144 151 furaibai 12 furaku 135 furakuSoN 135 furaneru 41, 86 furaNku 132 furaNneru 39,41,86 fiJraNSU 48, 140 furaQ8u 107 furasuko 33, 34 fureezu 106 fureNdo 146, 156 furil 149, 157, 205 furiidaiyaru 157 furiiraitaa 149 furilraNSU 149, 150 furiiraNsuarubaitaa 150 furilraNsuraitaa 149 furilsaizu 157 furiltaa 149, 150 furiQto 88 furoNto 148, 157 furoNtodesuku 148 furoNtogarasu 157 furooniNgeN 109 furu 1,141 furtlllCU 110 fuudo 87
Loanword index 245 fyoodoru 93 fyuujoN 181 fyuuzu 86,177 G gaadeN .p,, 105 gaaru 147 gaaze 51,52,99,129 gaido 156 gaidoraiN 57 gajeQto 97 gao iliNjeN 65 gaQCU 114 157 gaQcupoozu 157 garasu 28, 35, 38, 41, 42, 111, 157,168
gasoriN 157 gasoriNsutaNdo 157 gasu 28, 39, 128, 138 gasuru 138 geba 53 gebaruto 53 geecu 185 geemu 142, 157 geernuseNtaa 157 geemuseQto 142 geete 99, 172 geeto 185 geQtaa 157 geQto 55, 199 gereNde 52, 51 105 geru 53 gia 148 gipusu 39 girujia 68 giya 103 goete 172 gomu 'Qom' 84 gomu 'rubber' Lf, 38, 41, goo 155 googuru 82 goorudeN 156 goorudeNawaa 156 goorudo 55 goosutoQpu 155 gorira 149 gorufu 57 gosiQku 117 guaba 92 guadoruupu 121 gudanisuku 109 guQZu 110 gurabu 181
42
guramu 28, 49, 129, 187 guraNdo 101, 149 guraNdosutaNdo 149 guraNdosutaNdopuree 149 guraNpuri 49 gurasu 42 gurasunosua 46, 47 gurataN 48, 49 guraUNdo 101,149,157 guraUNdomanaa 157 gureborotaN 35, 53 guree 129 gureepu 153 guriiN 58, 106, 149, 157 guriiNhausu 149 guriiNkaa 157 guriru 141 guriseriN 124 guroi 139 guromuiko 98 guroobarizeesoN 154 guroobaru 205 guroobu 181 gurotesuku 139 gurujia 68 gurukoosu 185 guruupu 6o, 106 guuguru 138 gwoete 172 gyagu 75, 115, 141 gyagumaN 141 gyaNburu 98 gyaNgusutaa 91 gyara 98 gyararii 98 gyaru 156 gyoocu 172 gyooza 8, 62, 6:;, 64 gyuute 172
H haafu 156 haagu 88 haamonaizedoN 203, 204 haarnonii 138 haareedabiQdosoN 100 haato 141 haatofuru 141 hacaturiyaN 181 hai 140, 144 haitiiN 140 haiyaa 103 hamaa 90
hamaasu 87 hamoru 138, 145 hamu 38, 41, 87 hamusutaa 90 haNburuku 105 haNdo 144 haNdoru 154 hanemuuN 44 haNgaa 144 haNgaasutoraiki 91, 144 haNguru 67, 84 haNkaci 55, 124, 125 haNkaciifu 125 haNkeci 124, 125 haNkeciifu 125 haNkeruCiifu 124 haNma 176 haNmaa 90 haNsuto 144 145 haNtindoN 91 hanukaa 87 haQ(ibaqku 119 haqpii 58, 116, 141 haqpiieNdo 141 harasurneNto 144 harii 146 hariipoQtaa 146 haripota 146 harogeN 40 haromoni 146 haroo 146, 157 haroomooniNgu 146 haroowaaku 157 harowa 157 hasiQsu 109 hasiSi 109 hasisu 109 hausu 55, 149, 157 hea 103 heapiN 174 heboN 124 heNto 88 heQdiNgu 117 heQpubaaN 124 heQto 41 herupaa 199 herliSiNgubori 91 herusu 156 heya 103 heyapiN 174 hii 55 hiidesu 32 hiiriNgu 57
246 Loanword index hiiriyo 169 hiNduu 83 hiNzuu 83 hiQpare 146 hiQpu 178 hiQto 146, 178 hiQtopareedo 146 hire 49, 95, 103 hisuterii 99 hoeeruuoQCiNgu 93 hoiQsuru 93. nil homupe 147 hoN 157 boNkoN 64 hoodeN 51 hoomu 86, 135, 147, 198, 199 hoomuherupaa 199 hoomupeeji 147, 154 hooru 157 hoQWcisu 68 hoQk.ee So, 100 boQkii So hoQk.u 41 boQpu 41 hoQtodoQgu 97 horaa 87 horumoN 53, 107 hoteru 55, 59, 6o, 61, 129, 150, 156, 187
howaito 12, 93, 156 howai.todee 12,156 howaitosacu 124 hyuQte 52, 122 hyuuzu 86, 177 I
i myoNbak.u 65 ibeNto 200 ibu 106,149 ideorogii 52, 53 ieena 93 ierookaado 93 iesu 58 igirisu 33. 34> 68 ii 1, 8, 64 iisuto 93 iiyuu 189 ikooru 92 ik.ura 28,46,47,154 imeeji 114 iN 1, 6o, 141 iNborubumeJ:o.'to 203, 204 lNci 109,178
iNCoN 66 iNfoomudo 142, 205, 206 U."foomudok.oNseNto 142, ::w;, 206
U."fura 133, 195 lNfurasutorak.ueaa. 13:1 lNfure 13.3, 1..34, 1.35 l~"fureesoN 131135
iNheruno 32 inHaru n8 iNkyubee8oN 202, 203. 204 iN kyu.beetaa 201 inobeesoN 195 iNpo 52,133 iNpoteNCU. 52, 133 iNpureQSoN 106 iNseNtibu 57 iNsutaNto 57 iNsuto 136 iNsutooru 136 iNsutorakusoN 1.J(i iNsuto.ralmtaa 136 iNsutorurneNtaru 136 ii>;'taa 58 iNtaaneQto 194 iNtabyuu 177 iNtavyuu 177 lNterl 28, 46, 47, 133, 134 lNterlgeNea 131134 iNterlgeNCia 133, 134 iNI]faada 43 i!l.'tora 136 iNtoro 132 iraku 6o iraN 129 irumaN 32, 33, 170 itakaji 148 itariaN 148 i.tariaNkajuaru 148 ivaN 85 ivu 106 h'l'aN 85, 86 iyomaNte 66 izumu 1,141
jaNkufuudo .57 jaNk.usoN 106, 130 jaNru 49,50,87 japaniaN 150 japaniizu 150 japaniizuajiaN 150 jaQguru. n8 jaQji 114 jaQki 107 jaQk.u 107 jazu 56, 115, 174 jeeaaru 190 jeepoQpusu 119 jeeru 99 jeNtorwnaN 88 jeQto 75 jeQtoeNjiN 175 jerii 88 jesueaa 72, 175 ji 59, 60, 188 jibaN 34 jigitarisu 39, 8.2 jihaado 43 jii 84 jiiai 57 jiibuijii 82 jiiNZU 101, 144, 145 jiiNzupaNCU 144 jiipaN 144> 145 jiipu 57 jiiri 91 jiN 95 jiNku.~u 154
jiNteeze 99 jireNma 83, 175 jiruba 125, 126 jogiNgu 88 joojia 68 jooro 34 jubaN 28, 33, 34 juneebu 68, 122 juneevu 68, uz juuru 99 juusu 154, 177 juusukiNto 99
J jaajaameN 64 89 jaajaNmeN 64,89,90 jagatara 38, 171. jakaruta 172 jamaika 67 jamiNgu 90 jamu 90
K kaa 144157 kaade 135 kaadegaN 135 kabusu 84 kacurecu uo, 147 kacurecusaNdoiQCi 147
Loanword index 247 kacll8aNdo 147 kadomiumu 146 ka.iro 68 kajuaru 148 kakao 41 kamera 44 97, 144 kamereoN 41 kamucacuka no kamyu 98 kaN 'Caen' 90 kaN 'can' 40, 41, 168 ka:t>.fareNsu 97 ka~11rru 39,86,90,129
kaNniNgu 154 kaNnu 89,90 kaNpa 46, 13,3,134 kaNpania 46, 47, 133 kaNsuto 145 ka:Ntasu 92 kaNtera 33, 34, 35, 40 kai-i:orii 144 ka.nuu 129 kapitaN 34 kapuseru 51, 52, 97 kaQpa 33, 34 35,123,151,152, 168, 173 kaQpuru 50 kaQtiNgu 174 kaQto 97, 140 karaa 59, 187 karaNdaQsu 121 karaQto 39 ka.ree 100, 152 kareNdaa 104 ka.riesu 53 karubl 67, 84 ka.ruka 33, 3+ 152 karukeQto 150 ka.ruki 39, 107 karuSiumu 40, 150
karusiumubisukeQto 150 karuta 34, 35, 151, 152, 168 ka.rute 52, 53 karuteru 52, 102 karuto 57 kasuteera 33, 34 kasutera 33, 34 katarogu 97, 168, 171 kateeteru 39, 129 katekizurno 32 katoorika 170 katoriQku 33 kauNserlNgu 137
kauNslru 145 kauNtaa 145 kaUJ:>.i:aapaato 20,3, 204 kauNt:aasutoQpu 145 kaUJ:>.i:O 101 kawaado 104 kazafusutaN 68,109 keaNzu 104 kebeQku 121 kecaQpu 117 keeki 107, 174 keesu 6o, 182 keesubaikeesu 182 kernisutorii 154 keQto 152 keredo 32, 111 kerosiN 101 keruN 98 kii 157 kiisaN 67 kiiseN 67 kimujoNiru 65 kirnu oQkyuN 66 kimua 67, 84, 129 kina 39,170 kiQCiN 75, 106, 119 kiQZU 107 kiriSitaN 31, 32, 111, 129, 170 kirisito 32, 33 kirisuto 32> 33 kiro 49, 186 kiroguramu 49, 95, 186 kiromeetoru 28, 49, 187 kirowaQto 178 kirugisu 68 kiseru 30, 152 kiyaNpu 178 kobaruto 40 komeNto 58 komiNteruN 46 komiQSON 154 komyunike 174 komyunikeesoN 148 koNbinaato 46 koNbini 130, 131, 133, 139, 198,199
koNbinieNSU 130, 131, 133 koNbinieNsusutoa 130, 131, 133
lamdakutaa 145 koNdeNsaa 145 JamdeNSU 141 koNdeNsmnlruku 141
koNdoomu So kone 132 konekui!oN 132 koNfareNsu 97 koNhisaN 32 koNkuriito 145 koNkuuru 48, 49 koNma 41 koNmyunike 174 koNpa 155 koNpaato 132 koNpanii 145, 155 koNpasu 39 koNpeetoo 34 koNpooneNto 145 koNpuralaNsu 203, 204 koNpureQkusu 144> 145 koNpyuuta 104, 182, 184 koNpyuutaa 104, 145, 146, 148, 182, 184
koNsaato 145 koNsasu 145 koNseNsasu 200 koNseNto 'consent' 142,205,206
koNseNto 'plug' 132, 133 koNseNtoriQku 133 koNseNtoriQkupuragu 133 koNslroo 55 koNsome 49 koNsoociamu 194 203, 204 koNsuto 131 koNte 148 koNtesuto 145 koNtinyuiti 148 koNtorakutaa 145 koNtorooru 138, 144, 145, 146 konyaQku 49, 50 koohii I+ 35, 38, 40, 41, 86, 149> 168
koojiNcefu 94 kookusu 52 kooraN 84 kooryaN 8,64, 84 koosu 140, 157 kooto 'coat' 149 kooto 'faeces' 51 kopeNhaageN 68 kopii So, 117, 142, 156 kopiipeesuto 142 kopipe 142 kopiQpe 142 koQku 35, 40, 41
248 Loanword index koQpii So koQpu 34> 35, 123 koQtoN 119 koraboreesoN 195 korepoN 136 korera 38, :l9 koresupoNdeNsu 1.,36 kori.ma 98 koromubia 178 koromubiya 178 koroNbia 178 koroQke 112, 125 koroQkeQto nS, 125 korori :;8, 39 korudoba 85 koruima 98 kosumoporitaN 130 kosuto 141 kosutodaUN 141 kotaN 66 kuooku 92 kuootarii 92. kuQkii 82. kuraaku 1.22 kurabu 171, 178 kurakuSoN 68 kura&i.karu 117 kurasu 59 kureemu 154 kureepu 50, 149, 153 kureepu
kwootarii 92, 176 kyabecu 5.5 kyabeeji 55 kyabiN 98 kyadii 98 kyamera 44, 97 kyamuptl 178 kyaNdee 83 kyaNdii 83 kyaNpu 152, 178 kyaQCi 156, 157 kyaQcihoN 157 kyaQCikopii 156 kyaQsu 115 kyaria 98 kyariaa 98, 103 kyasutaa 149 kyu.ijiinu 89, 94 kyuroQto 97 kyuuaNdoee 189 M maaboodoofu 63, 64 maajaN 64, 89, 90, 168 maaku 14. 104. 108 maamoQto 38 maasutorihito 109 madamutaQsoo 142 mafi.N 119 magajiN 85 magun eCi.Qku. 141 magWieSi.urnu 40 mahomeQto 172 mai 1,58,60,140,198,199 maibWlinu 140 maihoomu 198, 199 maiku 57, 132, 156 maikuro 140 maikurohoN 57 maikuroke 156 maikurokeesoN 156 majeNta 88 maiiQku 149 majiQkupeN 149 makaroni 157 maka.roniuesutaN 157 makudo 133 makudonarudo 133 mama 58, 90 maN 69, 75, 141, 1.42, 157 manaa 157 manaamood.o 157 manee 100
manifesu:to 154 maNna 32., 33 maNpawaa 143 rnaQCi 171 maQha 122 rnaQku 1J3, 140 maQkuseeku. 13.3 maQpu 97 maQsu 142. maQsupoteto 142. marasoN 57, 87 maroN 101 maruci 140 marufuu 32,170 maruko poro 183 marukusu 140 marumero 34 maruseeyu 107 masu 148 masukoml 148 masukomyunikeesoN 148 masuto 38, 41 mastutzU 94 matisu 122 maUNteN 145 mayo 50,133 mayoneezu 50, 89, 133 mayoraa 141 mazaa 55, 144 mazaakoNpureQkusu 144 maza.koN 144 mazeNta 88 meado 147 medobeejefu 94 medoree 100 meekaa 6o meeku 149 meekuaQpu 149 meeN 18.5 rneeNraiN 75 meeru 14, 147 rneeruadoresu 147 meetoru 28, 49, 6o, 61, I.U, 178, 187, 188 mejaa 141 mejaasupuUN 141 mekusuto 189, 190 memo 138 memorii 148 memoriisutiQku 148 memoru 138 memosut:e 148 meNCi 55
Loanword lndex 249 meNCibooru 5.5 meNCU 62,64 meNte 135 meNtenaNsu 135 meQka 94 meQse 5:l meQseeji. 117 merikeN 124 meriQto 200 meriyasu 34> 35 meruheN 52, 87, 141 meruheNCiQku 141 mesena 48, 50, 194 mesu 'Metz' 122 mesu 'scalper 39 met:abo 133 metaborlqku 133 metaboriQku8iNdoroomu 133 metoronoomu 56 midoru 157 midoruhooru. 157 miira 31 34> 129, 184 miisa :12, 169 mini 140 miNsi 55 miNSU 55 miQeeru 118 miqkusu 141, 142 miQkusubejitaburu 142 miQkusuguriru 141 miri. 27, 28, 49, 132, 186 mirimeetoru 27, 28, 49,132, 178,186,187,188 mirosebiQCi 123 mirufllyu 48,50,98 miruku 141, 151 mirukueokoreeto 151 mirukuseeki 87, 175; see also
maQkuieeku mirukuSeeki 8;, 175; see also
maQkuseeku misa 32,33 mi§iN 149 misu 'error' 133 misu 'miss' 8, 145 misudo 147 misukoN 145 misukoNtesuto 145 misumaQCi 57 misutaa 147 misutaadoonacu 147 misuteek:u 133 mltoN 119
moare 111 mobo 147 modaN 147 modaNbooi 147 modaNgaaru 147 moderu. 59, 147, 187 moga 147 mohaNmado 172 mohi 39, 40, 136 moNkii So moNsutaa 144 moNtoriooru 68 moodo 157 mooniNgu 146 moorusu 102 moosu 102 mootaa 150 mootaaapaato 150 mopaato 150 moQbu 114 moruh.ine 39, 40, 86, 136, 170 mo.rumoQto 38, 40 mosu 190 mosuko 68 mosukomyuuru 68 mosukuva 85 mosukuwa 68, 85 mubeki 91 mUNbai 68 muQraa 73 muudii 154 muudo 156 myuN.heN 67 myuujiQku 149
N naburaCirowa 84 nacuraru 117 naiibu 154 naimeeheN 88 nairoN 150 naitaa 8 naji 84> 109 naNbaa 6o naNseNSU 58 napori 67 naqcu 110 naQSiNgtt 117 nataru 32, 33 nau 139 naui 139 Ndebere 91 neemu 178
neetibu 14!1 neetibusuplikaa 148 negurekuto 139 negurije 50 neguru 139 neQto 1J2, 146 neQtorajio 146 neQtowaaku 104 neru 39, 41 netoraji 146 niito 189 niizu 57, 89, no, 154 200 nikado 146 niQkado 146 niQkeru 40, 122, 146 niQkerukadomiwnu 146 nisu 135 noboSibirusuku 85 noirooze 52, 51 99, 100 noN 58,140 noNsekusoN 1.40 noo 58,140 noohau 57 nookaqto 140 nookomeNto 58 noomaraizeesoN 154 205 noosoopiNgu 99 nooto 149 nootobuqku 149 norwna 28, 46, 47, 91 novosibirusuku 85 nugaa 50 nyuruNbe.ruku 98 nyuu 156, 178 nyuufeesu 178 nyuuhaafu 156 nyuusu 57, So, 100, 149 nyuusukyasutaa 149 nyuutorino 140 nyuuyooku 6o, 172 nyuuyoruku 172 0 oafu 84 obu 12,13,60 obuN So ofisu 148, 201 ofisukoNpyuu.taa 148 ofu 58, 115, 128, 141, 142, 202, 203,204 ofukoN 148 ofureko 142 ofurekoodo 142
2.50
Loanword index ofusaido 97 ofusaitoseNtaa 202, 203, 204 ohoocuku 87 okutaN 144 omuraisu 48 omurecu 48, 49, so, no ON 141,142 onanii 53 onapeQto 156 oNea 142 ONIDON 67 onomatope 168 ONSU 39 ONZaroQku 141 oobaa 1, 140, 149, 173 oobaacua 103 oobaadoraQgu 140 oobaakooto 149 oobii 140, 189 oobUN So oodoburu 49, 98 ooeru 190 ookee ;8 ookusoN 106 oopuN 1o6, 140 oorai 58, 124 ooraioorai 124 ooru 'alf .58 ooru 'oar' 102 oorudodeQto 55 ooru
paareNteeze 134 paasonaru 146 paasonarukoNpyuutaa 146 paasueeto 55 paatere 169 paa.teru 169 paatii 72, 156 paato 149 paatotaimu 149 pabu 106 paburiQku 104> 203, 204 paburiQkuafeaazu 104 paburiQkuiNborubumeNto 203. 204 paburo 106 paeriya 91 pafu 149 pai 'mahjong tile' 64, 84 pai 'n' 75 paipu 155 pajama 141 palm 123 paN 28, :11, 32.65 paN gimuN 65 paNCU 107, 144 panikuru 138 paniQku 138 paniQkuru 138 paNsaa 89 papa 58 papu 123 paQCi 67 paragwai 181 pareedo 146 pari 67, 87, 178 pariigu 147 parodii 139 paroru 139
pa.SifiQkuriigu 147 pasokoN 146,198,199 pasukuwa 32 pasuta 43 patokaa 144, 145 patorooru 144 patoroorukaa 144 pauCi 100 paudaa 149 paudaapafu 149 pea 157 pearuQku 157 peci.k:a 46 pedaNCiQku 141
peehaa 188, 189 peeji 59, 109, 1.17, 168, 178, 185, 187,188 peeNto 175 peepaa 147, 149 peesuto 142 peN 149,157 peNdiNgu ;8, 153 peNihi 89 peNki 41, 107 peNsiru 69 peQtO 156 pera 135 peresutoroika 28, 46, 47 perumi 109 pesuto 39 peteN 63
piafu
122
piano 174 piasu 104 piiaaru 189 piieeCi 189 piikee 190 piimaN so, 129,155 pii~u 175 piitiiee 189 piNiN 64 piNku 91, 139, 156 piNkui 139 piNkumuudo 156 piNkusaroN ..56 piNpoN 125 piNto 29, 149 piQca 43 piQeaa 117 piroSiki 46 pisutoru 41, 153, 171 piyano 174 piza 43 poatie m poCiNgu 125 poiNto 59, 6o, 61, 149, 157 poiNtogeQtaa 157 pokeberu 69 pokemoN 144 pokeQto 144 pokeQtomoNsutaa 144 poNeaQku 67, 84 poNdo 35, 38, 39, 40, 41 poNpu 39, 171 )JONSU JS, 39 poNZll 38, 39 poo 83
Loanword index poocidoeQgu 142 pooku 104 pooraNdo 182 poozu 157 poQdeNgu 125 poQpusu 143 p.Jrisii 130 porisu 105 poroheeta 111, 169 porucugaru 34 poruno 102 JX>runoguratJ.i 102 porutogaru 34> 35 JX>SutO 56 potaaji 109 JX>taaj u 109 potefura 144 poteto 11, 12, 142, 144 potetofurai 144 puaaruea 64 pudiNgu 125 pufarucu 89 pufenihi 89 puQciNgu 125 puQjiNgu 125 puracina 40 puragu 133 puraibasii 57 puraido 91 puraisiNgu 203, 204 puraisu 141 puraisuofu 141 purarno 147 puraQtofoomu 86, 175 puraQtohoornu 135, 175 purasu 157 purasuciQku 147 purasuciQkurnoderu 147 purasudoraibaa 157 puree 55, 149, 156 pureegaido 156 purekoN 145 purekyasuto 145 purekyasutokoNkuriito 145 purezeN 134 purezeNtee8oN 134 purezeNto 59 puriN 125, 126, 129 puriNta 104 puriNtaa 104 puro 6o, 61, 136, 148 purodyuusaa 177 purofeiionaru 6o
purofiiru 148 purofu 148 puroguramu 6o puropagaNda 6o puropera 135 puropoozu 137 puroretariaato 6o puroseQsaa 144 purosucicuuto 6o purukogi 67, 84 pusaN 66 puuCiN 94 puuru 107 puuilikiN 93, 109 pyoNyaN 66, 84 90 pyuum a 93
R I'a1
raameN 64 168 rabaa 144 rabu us, 156 rabuho 156 rabuhoteru 156 raibu 157 raibuhausu 157 raiCii 64 raifusutairu 57 raiN 100 raipuciQhi 72 raisu 48, 154 raitaa 'lighter' 176 raitaa \vriter' 149 raito 157 rajikoN 145 rajio 8, 8o, 145, 146, 173 rajiokoNtorooru 145 rarnu 14, 142 rarnune 124 ramuuuru 142 raN 142 raNbiki 121 124 raNdo 'backpack' 41 raNdo 'land' 156 raNdoseru 41 ranerii 82 raNguUN 68 raNjerii 50 raNkiNgu 91 raNpu 41 raQkii 58, 142 raQkiisebuN 142 raQpa 38,41
rarii 129 rasa 33> 34 ratatuiyu 87 razaanya 91 razaniya 91 rebaa 157 reberu 141 reberuaQpu 141 rebyuu 148 redii 141 rediifaasuto 141 reeauto 182 reeN 101 reeniN 172 reeNkooto 35 reeNpa 39 reepu 185 reesu 'lace' 100 reesu 'race' 59 refu 109 regee 185 regiNSU 125 rejaa 57 reji 132 rejisutaa 132 rekoodaa 147 rekoodo 57, 142 rekurieeiloN 194 remoN 89, 147 remoneedo 124 remoNsukaQsu 147, 176 remoNsukwaQSu 176 reNji 101 reNtogeN 51, 52, 98 reNzu 28,39 repura 40, 42 reQsLL~
106, 119
reQteru 41, 122 resipieNto 202, 203, 204 resuka 147 resuteru 150 resutohoteru 150 resutooraN 126 resutooraNto 126 resutoraN 50, 126 resutoraNto 126 resuturaNto 126 retoruto 39 revyuu 181 rezaa 87 riboN 129 ribu 142 riciumu 40
251
2.52
Loanword index rihablrl 133,199 rihablriteesoN 133 rihiteNsutaiN 109 riibe 52 riigu 56, 147 riizonahuru 58 riizu 88 rikaa 117 rikuruuto 57 rimokoN 144, 145 rimooto 144 riNgu 12, 13,142 riNkaN 172 riNkoru.'i 172 riNpa 39, 40, 51 rinyuuaru 168 riQtoru 49, 122 risaikuru 138, 194 risuboN 68 risuku 200 risuriN 124 risuto 12 risutora 13:1 risutorakucuariNgu 133 riteraSii 202, 2o:t 204 rizooto 144 rizootorabaa 144 rizoraba 144 rizumu 107 roguiN 195 roke 156 rokee8oN 156 romaN 171 roNdoN 6;, 74 roNdonia 101 roNgu 140 roNrii 101 roodo 157, 203, 204 roodoobuzari.Ngu 12 roodopuraiSiNgu 203, 204 rooN 101, 154 roosu 55,124 roosuto 124 roQho 109 roQku 91, 141 rorlita 145 roriitakoNpureQkusu 145 rorikoN 145 rosuke 46 rosutoropooviQCi 123 rozario 32, 33 ruifiriQpu m ruipe 66
rtiNpeN 53 ruporutaaju •19, 105, 122 ruQku 156, 157 ruQkusu 97 ruucu 57, 110 rtllliiill 88 ruuru 91 ryazani 93 ryosa 91 ryuQku 52 !)'UQkusaQku 52, 129 ryuuju 99 f}'UumaCi 142 ryuuto 100
s saabeeraNsu 202, 203, 204 saabisu 6o, 149, 155, 156, 199,205 saafiN 125, 126 saatiN yuuesuee 126 saafiNgu 125 saakuraa 141 saakuru 141 saarnoN 141 sabaNna 89 saboN 35 saboru 48, 49, 1)8, 139 sabotaaju 138 saboteN 35 sa.iaia 103 safuraN 41 sa.ido 157, 203, 204 saidobureeki 157 saieNsu 154 saikuru 138 saiN 137, 156, 157 saiNpaatii 1.56 sa.iNpeN 157 saito 202, 203, 204 saizu 28, 59, 157, 187 sakarameNto 32, 111, 112 sakirihiiSiyo 169 sakusesu 156 sakusesuruQku 156 samaa 145 samaakoNsaato 145 samakoN 145 samowaaru 46, 129 saNdo 'sand' 149 saNdo 'sandwich' 133 sa.NdoiQCi 133, 147, 176 saNdopeepaa 149
saNdouiQCi 176 saNgurasu .57, 141 saNkyuu 58 saNpaQpu 123 saNpauro 101 saNto 31,32 saNtora 145 saNtosu 31 sapurai 203, 204 sapuraisaido 203, 204 sapuri 133 sapurime:l¥1:o 133 saQkaa 57, 97 saQkariN 116 saQparu 55 saQSi 109 saQsu 109 sarabureQdo 21, 114 sarabureQto 115 sarada 6o, 112, 182 saradabooru 182 sarami 43 sararii 141 sarariimaN 141 saroN 156 sauNdotoraQku 145 sausupoo 185 sebiria 85 sebiriya 67 sebUN 142 securumeNI:o 110 seebu 84 seedo 87 seefu 55 seemi 40, 42, 44, 87 seenu 89 seerusu 84> 149 seetaa 92 seeuCi 45, 46 sefure 146 sejoN 66 sekohaN 144 145 sekoNdo 144 sekoNdohaNdo 144 sekuhara 144 sekusii 57, 108 sekuSoN 140 sekuSuaru 144 sekuSuaruharaswneNto 144 semeN 125 semeNto 39, 125 semi. 140 semlroNgu 140
Loanword lndex 253 seNCi 28, .19, 133, 134 seNCimeetoru 28, 49, 133, 178, 187,188 seNSU 95 seNtaa 59, 6o, 61., 157, 202, 203,204 sepaado 87 seQkusu 108, 146, 156 seQkusufureNdo 146, 156 seQto 28, 58, 59, 142, 187 sero 8S seru 39,41 serufu 149 serufusaabisu 149 seruji 39, 41 sezaNnu 89 sli 84 siizUN S4 sobieQto 46, 93 sobieto 46, 93 sofutoaisukuriimu. 149 sofutokuriirnu 149 somurie 50 sooda 39 sooiNgu 149 sooiNgwniSi.N 149 soopu 156 soopuraNdo 156 soorufuru 185 soQpu 42 soruj eniiciN 181 sosuuru 99 subeta 34 suCiiru 149 sucuaato 104 suc~adesu 177 sucuwaado 104 sufu148 sll.imiNgu 90, 142 sll.iQcaa 75 sll.iQCi 92 sll.isu 140 sukaQ§u 92, 147 sukeboo 144 sukeeto 144 sukeetoboodo 144 sukeQCi 114 sukeruco 72 sukii 6o, 129, 141, 17S sukiimu 200 sukiN 141 stooNSiQpu 141. sukipoocu 87
sukoQCiteepu 97 sukoQptL 40, 41, 87 sukuea 176 sukura.Nburu. 141 sukuraNburu.eQgtt 1.41 sukuriiN 106 sukuryuu. 149 sukuryuudoraibaa 149 sukwea 176 sukyaN 157 sukyanaa S9 sukyaQbu 114 sumaato 155 sumisu 115 sumisurofu 98 sumooku 141 sumookusaamoN 141 sumooru 157 sumooruraito 157 sumoQgu 114 149 sumoQku 115 sumuujii 87 sunoQbu 114 supaa 99 supageQti 43 supaiku 157 supaikutaiya 157 supakoN 146 supeeN 68 superu 141 supiikaa 148 supL""l 129 supiricu 32, no supiricuaru 31 supoito 39, 100 supoNji So supoNsaa 57 supoocu 59, 6o, 61,103,187 supoocuweaa 103 supookusumaN 10S supuriQca 117 supuu.>i 101, 141 supuutoniku 46 suraQgaa 117 surii 58, 141 suriisutoraiki 141 surirnu 140 suriQpa 176 suriraa 106 suroogaN 56 suroQtoru 118 sutaa 178, 184 sutaabaQkusu 138
sutaadasuto 148 sutaadasutorebyuu 148 sutaato 6o sutabaru 138, 145 sutadio So sutairiQliu 115 sutairu 59, 1.45 sutairukaUNsiru 145 sutajiamu 176 sutajio So sutakaN 145 sutaNdo 157 sutaNdopuree 149 sutaQdoresu. 149 sutaQdoresutaiya 149 sutaQfu 73, 115, 194 sutarebi 148 suteeki 107 suteepuru 148 suteepurufaibaa 14S suteetasu 107 suteNresu 101, 149 suteNresusuCii.ru 149 suteQkaa 83 suteQki S3, 107, 156, 171 suteQkigyaru 156 sutiQku 107, 148 suto 133, 145 sutoa 130, 131, 133 sutoobu 155 sutooru 144 sutoQku 28 sutoQpu 59, 145, 151, 155, 156 sutoqpuuoQCi 176 sutoQpuwoQCi 176 sutoragttru 118 sutoraiki SS, 107, 133, 141, 144 sutoraiku 57, 107 sutoraQguru 11S sutoreeto 110 sutoresu 57, 194 sutorikiniine 40, 111, 112 sutoriqpwoo 57 sutoroo 99 sutyudio So su~adesu 177 suucu .56,110 suucukeesu 110 suupaa 59, 140, 146, 14S, 1S7 suupaakoNpyuutaa 146 suupaamaakeQto 59, 14S suupu 42 suweedeN 67, 1S1
254 Loanword index
s saapeN 69 saapu 69 saapupeNsiru 69 saboN 33> 35 sacu 55, no, n6 saN 52 saNce 52, 53, 72, 181 saneraa 141 saneru 141 saNhai 64, 87 SaNpaN 49, 50 saNpeN 50 saNpuu 87,142 saNpuuseQtO 142 saQpo 48, 49> 50, 97> 122 saQtaa 156, 174 saQtaacaNSU 156 sarutoruuzu 99 satoru n8 sawaa 103 seawea 103 sebiija 81 seedo 87, 175 seekaa 181 seeku 133 seekusupla 172 sefu 72 sepaado 87 siaazu 104 siataa 87 sibia 139 sii 84 siicu 110 siidii 189 siiemu 190 siiN 85, 99 siito 110 siizUN 84, 141 siiZUNiN 141 simupojiumu 174> 182 simyureesoN 205 SiNboru 90 siNdoroomu 133 siNgurusu 142 siNkuronaizudo 142 siNkuronaizudosuimiNgu 142 siNnaa 89 SiNnyoro 34 35 SiNpo 134 SiNpojiurnu 105, 134, 182 siQku 49 siQpu 141
siriizu 59 sirikoNbaree 100 siroQpu 41 sirubaa 140, 155 siSamo 66 sisukoN 145 sisutemu 59, 60, 61, 145> 187,201 sisutemukoNpooneNto 145 siyaa 103 SoO 100, 156, 178 &oobooto 178 &ooroNpoo 64 84 &osutakoviQCi 109 sunaQpusu 122 ilunookeru 53 supuuru 168 SUQCU 99 suraafu 53 suraafuzaQku 52, 53 surafu 53 surafuzaqku 53 suurnai 64 suuzu 143 suwarucukoQpu 107
T taataN 82 tabako 28, 33, 34, 35, 140, 168, 170, 173, 178 tacuu 82 tai 60 taiga 46 tairnu 149, 156 taimusaabisu 156 taipee 64 taipu 28, 58, 59, 18 7 taiya 149, 157, 174 takisii 107, 1os, 182 takosu 143 taku~ii 107, 108, 182 taN 125 taNkini 149 taQkuru n8 taQkusufurii 119 taraQpu 111 tareNto 149 tarumomeetoru 39 tasukufoosu 194 203, 204 tatuu 72,82 teerna 28, 52, 99, 194 teepu 147 teepurekoodaa 147
teeze 53, 105 tegu 84 teguQki 97 teheraN 172 tekisasu 108 tekisuto 107, 108, 176 tekoNdoo 67, 84 tekusuto 107, 108 teN 157 teNkii 157 teNpura 28,33,34,35,168 tere 12, 28, 58, 59, 157, 187,188 terebi 59, 6o, 61, 133> 149> 186 terebijoN 59, 6o, 61, 133 terebitareNto 149 terefoN 28, 59 tereko 147 teresukyaN 157 tero 6o, 61, 134 terorizumu 6o, 61, 134 tii 154175 tiibiiesu 190 tiirnu 83, 175 tiiN 140 tiramisu 82 toeNtii 111 toerubu 111 toguru 118 toimeN 64 toireQtopeepaa 100 toku 123 toN 39 tonakai 66, 129 toobaNjaN 64, 84 toocika 46 tooiQku 189 topiQkusu 190 toQku 123 toQpu 6o, 146 toQpum 143 toraburu 138 toraiaNguru 135 toraNpu 154 toraNsupeareNsii 136 torapeN 136 torasuto 175 toreesabiritii 201, 203, 204 toreesabiritiisisutemu 201 toreesiNgu 147 toreesiNgupeepaa 147 torepe 147
Loanword index 255 toriQperu 51 toro 112 toroika 46 torornuso 98 toroNpuruiyu 98 toroQko 112 torusoQ 123 tosu 157 tosubaQtiNgu 157 totaN 35 towaraa 110 towaru 110 tuuruuzu 83 tyuuba 181
u ueebu 176 ueederUN 85 ueetaa 92 uefaasu 143 uesu u5 uesutaN 157 uesuto 125, 178 uiikuneQsu 55 uirusu 53 uisuke 172 uisuki 172 uisukii 172, 176 uocuka 46,47 uootaaporo 92 UOQka 28, 46, 47> 85 uQci 92,109 urajiosutoku 85 uretaN 28, 53, 97 usuke 172 usutaasoosu 92 umnaN 142 umnaNribu 142 uuroNca 63, 64 uuru 142 mvisuke 172 mvisukii 172
v vaioriN 85, 183 vaisubiiru 85 vaweNsa 101 venecia 73, 85 verudaN 101 verusaiyu 181 viiN 85 vitamiN 133, 175
vokyaburarii 175 vomiciN 172 vorufusuburuku 85
w waado 144 waadopuroseQsaa 144 waaku 157 waapuro 144, 145 wagunaa 105 waido 156 waidosoo 156 waiN 92 waisacu 55,124 waisubiiru 85 waiya 103 waka 3+ 35 wakuciN 51, 52, 88 waN 1, 58 waNerudiikee 189,190 wanisu 135 waNtaN 64, 84 92, 97 waQfuru 118 waru~awa 67 wa.SiNtoN 6o, 91 webu 75 webusaito 195 weebu 176 weeruzu 92 wesuto 178 wesutomiNsutaa 130 wicita 92 wiikueNdo 92, 154 wiiN 85 wikipedia 79, 132, 138 wikiru 138 winipegu 75 wisuki 172, 176 Vtisukii 172, 176 vli.sukoNsiN 97 WON 67,92 wooku 69, 75, 157 wookumaN 69, 75, 157 wootaa 156 wootaagyaru 156 woQka 28, 47, 85 wosureQto 69 wotoka 46, 47 wyuruteNberuku 177
y yaado 93, 125
yaaru 125 yaarudo 123, 125 yamuea 64 yaNgoN 68 yeecu 181 yeena 93 yeeru 76 yojimnu 168 yoNdaburyuudii 189,190 yooderu 93 yooguruto 102 yoQteru 150 yoQto 150 yoQtohoteru 150 yunesuko 189 yuniiku 89 yuQke 67, 93, 98 yuriika 184 yutorehito 98 vuu 1, 6o, 99, 134 yuufoo 188, 189 yuukari 133, 134 yuukariputasu 133 yuutopia 150
z za 1, 12, 13, ;8, 59, 6o, 75,140, 14}, 188 zaaruraNto 105 zairu 52, 53, 84 zemi 52,132 zeminaaru 52, 132 zeNCiyo 32 zeneba 68 zenekoN 145 zeneraru 88, 145 zenerarukoNtorakutaa 145 zenerarusutoraiki 88 ze~1:orumaN 88, 175 zeQkeN 52 zeraCiN 88 zeraniumu 88 zerii 88 zero So, 203,204 zeroemiQsoN 203, 204 zeroQkusu 139 zeroru 139 zesucaa 175 zii 84 zuboN 48, 49, 140 ZUQku 38, 4}, 82, 175 zuroosu no, 175
256
Loanword index 4 .;/7 1) :L
-=r-{ -
(4WD) 189, 190 140,141 7 7-7Jv 39, 170 7-1 6o 7-17/ 104 7 .-( -:../- (I C) 190 7 .-( .A.-71) -J, 100,149 7 .-( ;<, ·'7 1) ./ 100 7-1.A.l'\-/ 53 7 .-( ~7 ,., 108 7 -1-1:!/ 52,53 7-17 ..(-(IT) 189 7.-(f!ljyij 142 7 .-( }< Jl· 144
7 .-(;;. 67 7 '7 1-- 55. 57. 58, 156, 157 7'7 ]--::r-7, 157 7'7 H~-y-7;<, 156 7'7 1--Y-::.•yij 142,201 7fJ?Y:~ ~I) 7--r-·- 57,200,
20J, 204
7777/Q" 107 7 7-1!-lf I)- 174 77~~~1)7
202,
203,204
7·'7-l:!·;rijl)- 174 7 7 t!Jv 133, 134. 150 7'7-l:!kv-?r- 133.134 7';/7/ 150 7' ::; .: . :/ ~· 203, .204
7 :<. 7· 1) '7 :<.
7 ."Z.. t·U 77.1--l) 77,:~7
122.
~/~Y 125 ./~/1-- 125
133
7 .Al~7 jj';<, 133 7 7,1~7::¥'/ 52 7.A-"'.A 1-- 40 77-v'f"'Y7 141 7 'Y '"I 58, 6o, 137, 141, 146,149 7·y'"'IQ"v-¥ 119 7 'Y ,.fv 118, 154 7 'Y /Jv~:,; 1-- 119 7•:.:7- 73 77-J-?7- 105 7¥~'Y~s/ 117 7 )-I) .r.. 49 7 J!J...·."Z.. 147 77 189 77-of-.A 1-- 56 7 -j· '7 :,;·lj"· 57, 144 7·}·>'';7',./+)"-7 .-( l~ j]l 144
77 '7 /7. 137. 138 77 ~)]; 144 770'.1 115 7::.."'?' 32 7=-.:J. 133 7=-.:J.-~s / 133 7 ~~-~-- 134,150 7 ~~-... r :J.:,; l- 134 77:x:.7-:t:. 104 77?r-'l;t-~7. 149 77 :9-t!-Jv7.'1;l'- ~7. 149 77"'?'-0r'JY 109 77"'?'7-{-;!f...':/->r- ~ 107 7/1) /.1"1) 43 7""'•yl.7 50 7·"\"'?'117 32 7""'-~ 62 7~-{/ 1-- 132 7:J.71-- 146 7 :J. 1) t.J 140 7 j. ~) t.J >" 124 146, 149 7:J. 1) :b/::r-1::::- 149 7 :J. ~) t.J>'7·Y r :;f--;v 146 77-b. 150 77t.J;H 48,49 71)7.7./lJ./'.17 84
70 '.!7 128 7 Jl· t.J I) 39, 128, 140 7Jv:J-Jv 39 7 iv ;r ;J; 32 7 Jvl'\-1 ;r- 150 7 iv' "·1' r 28, 52, 135
7 Jv l '\1,. 176 7Jk\/)"l) 7.-7 109 7 fv/7. 56, 142 7Jt·~=-? b. 176 7 Jv~=-:L -b. 176 7Jt·:J.=-7 68 7!-·;];.:¥'- .28, 52 7'7- 156 7-:-lt.J- 104 7 :.-·'IJv 135 7 :_dr- 1-- 49, 121 7/::J-j]; 49, 107 7/':/g 32
7/7--r-7'
122.
7>'7Jvvl:::. 1-- 109 7/~ry l--0'7 98 7:-/ J!;>J-7 1, 58,104 7 / ~o-1 F 110 7'./ 1-- '7-/ 68 7 / '" 1) -!>::/.1].· .58 7:/~~s :/ 117
.1/ 132 ·.:c=-.7 90 -1 • ~s/l'\1.7 65 .-(- 1, 8, 64 .{-7. }- 93 -1 -.:::z.- (EU) 189 .-( r'j 106 .-( r'j'y:,; 85 -1 :x:.-'Y 181 .-( :x:.-7 93 -{::r.-7 93 .-( :x:.-J]; 76 .-( ::r.7. 58 -1 ::r.o-t.J- ~ 93 .-(.:¥'I} 7. 33. 34. 68 .-( 1.77 28, 46, 47> 154 -{:J-.Jl• 92 .-( :J.,'b. 1, 141 -1?rt.J-;J148 .-( :9 I) 7 :;..· 148
.-(?l1J7/t.JY::I.71t• 148 ·1'7~·o.:¥'- 52,53 -1 ::::.~ ->t It• 118 ·1' /•'\-~s >' 195 -17 106,149 -1""'/ }- 200 .{ j. -9 114 -1::1"«/7 66 -171.7 6o -17>" 129 -1 Jv"'< ::.-- 32, 33. 170 .-{ '7 >" 85, 86 ·1 / 1, 6o, 141
.-(/of- ::I.-"'-~ g /
204 .-r:.-•of-::I.-"'-:9-1/7.:9/ f- 57 .-( :;..• 7. f- 136 -{/7. }--;]; 136
202, 203, .201
-1>'7.1--7·'7~3 / 136 -1/7.1-7-7:9-136 -1:,; 7. t-1t· :J. >.;; Jv 136 .-( /-!:! -:-1 7-{ 7 57 .-( >":9- 58 -{:,1;9-;!f...-y f- 194 .-( y :9 r'j ::I.-· 177
-1/:9~::1.-177
.-( /'!' 109, 178 -1>'f"s /
66
-{:,l"f-.(77-.;t 43 -1 / 7 I} 28, 46, 47, 133, 134 .{:,l"j-l)lj/7·7 13J.l34 •( :;.,• 7- I) lj :1'J- '>t 133, 134
Loanword Index 257 7-:t-?r- 156
-1>'1-7136 .{/l-0132
'?>t-!l-.:¥'-\?.lt' 156
-1>''-:t-L !-! 1.p.,205,206
');:>]--9-
-1>'7:t-b.~:::J/-!r/
t'J>t:f1'"?'71J- 175 ?:tv.:z.v'Yr 69 ':7 ;f 'Y ;I:; 28, 47, 85 ') ;:>]-,_y :t.J 2S, 46, 47. S5 ') /.l"~Y ;1:; 46, 47 '7-ir 1-JJ 46,47 .;;·~ J117 A. :1 Jv~7 85 '7-ir./ 67,92 7:7..7172 '7:7..?1-'.J-::Z. 92 7 'Y 7- 92, 109 '? .:z.Jl,'TY-"\Jt'7 177 779/.l"A. 1-7 s5 ?v?r>' 28,53.97 ')"f. A.:f-·" 172 ')Jf-:7..7172 ~/7 ~v:' 172 :r.7 142 ::r...{ ;t.' (AIDS) SS, 1S9 :r..-.:r..';( .:r.- (ANA) 1S9 .::r..-~.:::./1- ss :r..-7-(-.:r.L (ATM) 1S9 ::r..-'T;v 39 :r..(AP) 1S9 :r..-:1-1 (AV) 189 :C.-/[; 154 :r..;t;7 1J >:1;v7 93 .:r.::f-lj" -1'7 ..( >' '}' loS .:r..:f ,":Z. 28, 39. lOS, 176 .:r.::f :7..·77 Y -v s >'
l- l.P., zos, 206 -1>'77 133.195 -1>'77:7.. l-777-1'"- 133 ...( >' 7 ]/ 133. 134, 135 ·1' ";,-' 7 !-·-~' 3 >' 133,135 -{Y:fL/·YVs>' 106 ...( >'"')\l / 32 .{Y'7!.'7 >''Y sz, 133 -1'YJ.f-;v:1;>( >' 1- 201204 t'J7 ...( ::t I) >' 85, 183 t'f7-1' A.~-Jv 85 t'J7 1) .:I: >'-lJ' 101 '7-r-7.r./~ 92,154 '7~(·..·-?'.:f.'YA. 55 7-r·..··:/ 85 rJ-r->' 85 '7-r::f~'f-17 79,132,138 ') ..( :f-& 138 '7-1:7..4'-172 ')...( A.:f- 172,176 '7-1':7..7 172 '7-1 A.:::J/VY 97 t'J-( ?r ~ > 133· 175 ') ..( 'f-?r 92 '7-r::::..~'}' 75 '7-1 Jv'A 53 ')-"'?' >' 142 ')-"'?' >' 1) ?'142 '7-.IL· 142 ')-p >'7-1'" 6J, 64 ').:r.-?r- 92 ')I.-7.11·:'--' 85 ?.:r:-7· 176 '):...-:1176 '7 .:I: -.Jt• 92 ').:r.:;t., 125 •7 :r..7...; >' 157 >).:I: ;l, l- 178 •7 :r..7.. }· 125, 178 ?.:r:7.l- ~";,-'A.?r- 130 "} .:r:.:f. ''J ..( 7 73,85 ?.:r:?' 75 ')I.77-A. 143 ').:I: 7•-tt.-r l- 195 t'J.:r: ;v-tt -1 .::z. 181 t'J.:r:.IL·?f>' 101 ?:t--7 69,75,157 ') ;f --7 "'?' >' 69. 75. 157
:t:
e:-
"'?'-··7
to8
:r.:fA. t-77 1- 39 X.:f A.!~- i- 108
:C.7A. U2 X-7:7..77.:>(-Vs>' ""?'--?'loS .:r..-77;7 133 .:I:.·7 :7..'7/$-'s >" 133 .:r..~r v~/7 J2, 111, 112 .:I..:::J 72. 132, 13.5 :r..:l' 43 :r..:::J p -;/- 132, 135 :r..;<.::t-.:r.A. (SOS) 189 :r...::ziJ.r'\-V .:z. Ul I.A.'T 133 .:I:.;<. '7'7 -1 'Y 7 121, 133 I.A. :tv 'Y '.) 41 94 .:r..7-7 ··y 1- 48, 49, 50 I. ·Y Jf 141, 142 X.•y -7 ;<.(X) 14
.::r.. 'Y -7 :;t.,.r.;v (XL) 190 .:r. 'Y7- 191 .r.·y! .:r:Jv tn :r..::::..v?f 34. us .r.::::..:;t.,?( 34 :r..;< 89 .r.;<.{-V- (NEC) 190 .:r. ;< (NG) 1s9 :C.)l-;/-;t- (NGO) 1S9 :r.;<7.1·-7-r- (NTT) 190 .:r..':f. 1- ~7 150 :r...:f.;v¥-· 52, 140, 150 .:r..:'l:.;v¥-.::z.-J- ~7 150 :r..r~-v.:r·-r 69
v-
:C.!~- v-7-1' $.11'"-:f""~/V
Jv 69 .:r.~¥7
12
:I.. 7 I.),;, (FM) 1S9 .:r.:to :'-' 15 4 .:r..~ ·y~·s
>'
203,204
:r..L 12,13.55 .J:I) 'f->' S8 :r..,v::::..-::::..s 91 .:r.v:f 39,136 :r.. v:f7 lJ v7· 39 .:r. v:f7 39, 132, 136 .:r.t47.1vf!-IJ7-r- 136 .:r. Y'-?1- 104. 176 .:r.ov' 139 .:r.o=f"•y 7 139,141 :r.. '7 >'it 1) .3 32 .:r./7 1- m :r.>'Y-V 142 .:r.>'Y-~:/IJ :-,.1/;f 142 .:I../V/ 28, 59, t+t. 157, 187 I.YV/":Z.l--Jv 144 :I../ A.!- 144 I./:7..""'-::r S7 .::r../?r-7-1>'>( >' !- 136 .:r.::./.; >( 136 .:r. :' F 141, 142 :r.::..'l~7:/ 142 I. :;.1,1 ~ ':l· ... j. :;.1 }· 194> Z02, 203, 204, 206 I./7 ;f-A.;>(:'-' 1- 194> 203, 204 ;t77 S4 ::t-t'J:,-' So ;t-.:r.;\.• (OL) 190 ;t-7V:s >' 106 ;t-7- 58 ;t-A. !-7" 7 140 ;t-'.J IJ7-{- 173
;v
;v-
258
Loanword index ;t-- !-' ?' jt; 49, 98
:'~ 140 ;.t·- t-:'.'- (OB) 140, 189 :<~"-?'::./ 80 :
:<1"7-.'f-7 32 :
;b l) :r.;z 53
i:J-7 135 7.7-.::;tj:j:/ 135 jj-::;t~ 42· 105
:h I) ?t 34> 35. 151, 152, 168 7:J JL·:h 31 34> 152
n ;v:::f-
jj-)]., 147
jj.{
¥ 156
:ii-1 ~7-1':.-' 57 ;b.{o 68 :h7>'VJv 145 :h? :/-l! I):/~ 137 :h7Y?t- 145 :h?Y?t-1~-1-
201204
t.J>JY?tA 1-'.Y/ 145 ;b'? :/ 1- 101
t.J:
39, 107
7:JJL·'T•y 1- 150 :h ivv 7 L. 40, 150
'V:/-;/' J:.~ 65
;b;b ;t 41
7:J+J7 A ?t :/ 68, 109
JJY: '->' 1- 97 i:JY :r. 7 Jv 148 iJ'.:Z. 28, 39, 128, 138 t.J A 7-7 31 34 7.17-.77 33.34 t.J.A~ 138
;b!V'~'?L,~.A'T'Y 1- 150 tJJVr 52,53 ;b JV'r JV 52, 102 :hJH 57 ;b Jv ~ 67, 84 tJv- 100,152 nt-->·71'- 104 7:J '7- }t 104 :h ~ 40, 41, 90, 168 f.J>'A 1- 145 ;b;;,t.;t;z., 92 :h >'77 33. 34> 35, 40 :h :/ 1- I) - 144 ;b >'=-~ ~ 154 :h >''){ 89,90 7:1~1~ 46,133,134 ;b~l~=-.7 46, 47> 133 7:1/~ :rt,:-·;z 97 n ~7 Jv 39, 86, 90, 129
:<)- /"'"<' f-.A( 168
jj'.} ~) :/
:<1"7 58, 115, 128, 14], 142. 202,
jj',.l I) ./ ;z, ?t ~ ¥ 157 jJ ?t 0 ij 97. 168, 171 ;b';l-1}- :/ !-' 147
:¥'7148
J:J '.Y 'Y 114, 157
::f- 157
?.l"7-( .7-.::I~~.::l.?t-148
7.5'.>' y;f--;;( 157
::f-·~t~
:t'7:.1~148
:h •;:' 1- 97, 140
.'i:J ·.y 1 ~ 31 34> 35· 123,151, 152, 168,173 .'i:J .y / Jv .5o :h'Y v'Y 110,147 7J'JVY-IJ">''r-(•y7- 147 JJ7-7 Jv 39, 129 7J7:f .A'.:c- 32 jJ !- ? l) .iJ 170 ;t; ~ .:;, 7 L. 146 :tJ 1- ~)'}' 7 33 :h~- 129 :tJ~?t~ 34 JJ7A 84 :h:t-l!;v 51, sz, 97 ;b~:r. 98 :h.L.7'-4r'YJJ 110 ;b :;l. 7 44> 97, 144 t.J:;l. v
203,204
:<J-:1 12, 13, 6o ;d-7 -r A 148, 201
J< 97
;t 7 -lj .{ 1- {! )' ?t- 202, 20J, 204
:
80
:<J-.A\7 43 :t-Jt-'J-7 87 : 35, JS, 68 ;t l) V-Tll.' 6o :<)-;];:f-;]; JS, 4], 88 *~
w. 142
:
:
67
:h- 144> 157 ;tJ-~ 51, .52, 99.129
157
JJ'.Y7-(:/~174
67 67 ::ftuL· 30, 152 :::f- 'Y ::f. 107 ::f 'Y f· '..-' 75, 106, 119 :::f-·r 39,170 :¥'"fA 39 ::f-·1!~
:fl.. :t'.Y::f:r. ~ 66 :fl.· Ys ~.{ f[; 65 ::fL. 7- 67, 84, 129
:¥'-'f 103 :¥' 4[~ 75,115,141 :¥'1"~~ 141 ::f "\'.A .;r- 149 ::f 1" '.Y :::-·· :r. 115 ::f "\' '.Y 7- 156, 157 ::f-1"'.>'7'-::I~- 156 ::f4fy7-Jt~ 157 ::f 4[ 7-( - 98 ::f4r ~:/ 98 ::f 4[.A\-9 55 ::f .;-.A(·y 55 ::f 4[ L. / 178 :;:f-.;- :;l. 7 44> 97 .:¥'4r? 98 ¥-\•7 1) ·.... 98
Loanword lndex 259 ~-~·I)
7 98
'*"'"1) 7,.____
7'7 :.-· FA 7 :.-· V 149 77Y F::z_.y.y F7'
98, 1.0]
v- 149
¥-~· ;]; 156 -~-\' 7'A :)1·--- 91
>
Jf'.'l :5<·---- 157 !- 152 'f·y 1-- 55,199 ·J"t'\ 53 'f)<.) [.• ]-. .53 ·]· ''!
~ -~· '.;,I ·i"·1' 83 ~--1'>7"- 83
7'7Y7' 1J 49 7 1)7 103 7' !J - Y ss, 106, 149, 157
=+' -'t :,:.;:;t 152, 178
ij'l)->;l.J- 157
7
~--\">"/ 178 ¥.,. y·-r )]; 98
~li)·--·';/J\";i;:z, 149
lf ;J; 53
of- ;:t .-( :/ ·--- ;<
:.1r-::l-7j~::c-
(Q&A) 189
·::f::l-7- 172 of-::2. P'Y t- 97 .:f g -"! 8, 62, 63, 64 .:fg -'Y 172 .:} 1) v -7 y 3J, 32, 111, 129,170 .-;f. i) :.-· I- 32, 33 ~ ') A }· .32, lJ ·*iii~;z, 68 ¥;k: 77 68 4~ t:I 49, 186 =j.: 0 ~f7 b 49> 95> 186 ~-0)-}-;J; 28, 49> 187 =+' 0 ';7 'Y t- 178 77 F /I;-/ 121
:J >
Y"/-~i
~;!'--;>'.@ 138 7·····r7·--- 49 7- 1) y 7;t7 153 ·7 .::c."A.7'·s -:-l""<'-7
98
32,111
!;Jl) A'"' A 33, 145, 149 7 1) A""<' ::z_.,( :1 149 -71) 7""?' A :::I ';/J~-=--- 145 -71) A""<' .7. 'Y 1)- 110 ~:J' 1) {r 1) >" 124 7 1) 'Y !;J 114 71J Jv 141 7 ;];7·---'/ 84 !;J;l··---- 129 7Jk--·:::<' 107 'l/lJ-7 6o, 1o6 :7';J;:J- ;z, 185
.YV'/7" 52, 53,105
'l/li:/7' 68 ·7JVA 32
7'J.r-- 129 -7 v-1' so, 149, 153
7:t I.'T
7:t -j1 I)-
92,176 ~7:t-7 1)- 92 7'Y:::.::z. -7 109 7·y:t-·- 82 7·y ;;(' 110 122
~ 101, 149> 157 ~'!7 t) -:-/ ~""?'7-- 157 ;,-;7~7::-·s
~77
:J7Y 66
139
-li-Vs Y 154
172
·7:t-7 92
77---7 Jf-7 •7 /
~ ::,; 35' 53
:1o-;~ 1J
Y 68
v f.J /[; 117
7·;7 .7..
59 7";-;A 42
7'7 A/ .A 1- 46, 47 7';;-;.y.y 48,49
777' 171, 178 77:1181 /f:;7 L.. 28, 49, 129, 187 Jf7 >" :~ 101) 149
144
:::I-'f-7~5>132
!;JOA:i:J)' )--- 1) - 144 :107-A-7 139
'lob-{:J 98 ·700:;tJL.. b _52 ~~ 0 r; ':;! +) :,! 48, 49, 77>'7.' 104 lf-(''J 185 7---.f' 107, 174 7-----;:z, 60,182
7-A.
~~-{.
'f-1- 185 -t-1., 142, 157 'f-L. f'c 'Y 1- 142 -1-L·t:Y 7- 157 7'1--'r,»y/ 117
::JJ":Jt.•' 40 :J ~- So, 117,142,156 ::J~-o/{_-;:z,}-
111
·7-A
-t.:L.-5'- 99, 172
::I 'Y !;J 35> 40, 41 :J 'Y 1-- y 119 ::r '7 ~ . .-. So :J '.'/ / 34, 35, 123 :::I:::..->t 'Y ·7 49,50 :J-'f.-132
!;J 0;!:) > 144, 145
7DA
:J -::.7 > ~y I . "7 94 :J- A 140, 157 :I-A}· 'Y 1' 155 :::r·---l- 51, 149 :J ·----!:: ·---- 14, 35, 38, 40, 41, 86, 149,168 :J-7-:-1 84 :J- 1) -\•Y 8, 64,84 :j'-Jt.·-7"> 156 ::l·----;J;·;F''>7 '7 ·---- 156
~JA t- 141 :::I A }· .j'? Y 141 ::1 ;z-"C;J~ 1J 5- > 130
149
7o7'f·7 67 -j1'Q .,...
?"PVY 101 :i- 155 :J-!;7 A 52 :i-'l;J; 82
:f-/V "f' 55 ::j'::_,·,y 7 117
7']._,.------f 153
·.7'l..- ;J~ p 181
154
7 ;];:·/
7J.r--L, 154
I- 92
}- 1) -
7-vr
145
-7v-:t'f':,;j 7-1
~A
7~!;:<,1-.or.•Y 33
!;Jl) 89' 94
--~]"./'\'.') ~7 121
142
:::I ~'yv{ 142 ::I~---{ 142
:::r.r\Yr..·----:TY 68 ~.1 ~ •y ~ 5
>
154
:::I~.::J.-=-·7 174
182
~.1 ~ :;2. ..::. ·7-Y 5 Y 148 :J.:::.Y-5'-.11.·>' 46
~j'£, 14> JS, 41, 42, 84 :J) y }- 58 ::I';7;)q_.-~/S >' 19.5 :J I)""?' 98 ::f IJ 7 149
26o
Loanword index :::J.Tl·-1'~ :::JJld~/~
98 85
:f.TV"i 57 :::~t...-;:.J.f-:--7:--;:. 136 :::1 1,.-;f-y 136 ::Jt..--'7 38,39 ::I 0 'Y '7 112, 125 ::10 ~'l7'Y ]- 118,125 :JOL.~7 178 :JOb.~{' 178 :::10 IJ 38,39 :::10/~7 178 :f~7-1j'2 :::~:--7-;Lr
48,49
::l/'71) -1- 145
:::J:..i'ij--J- 145 :l/V""A 145 :::1:.-.'VO- 55 :::1/~"P- 55 :::1/.Al- 131 :::C/f::y-!j-_7, 200 :::1 / {; / 1- 132, 133,142, 205,206 :J>t:.-1 1- 1) 'Y~1 133 :::l/{0 J-l) 'Y-'7/7;/ 133 :::I/'}-'f-7 L. 194>203,204 :::l/';1;>( 49 :::I/?''7!1- 145 :::J/71{8
:::r 'Y'T ..( :::..::v( 'f· .{ 148 ::I/7~.A.!- 145 ::J>'f"y'ij-- 145 :::1 >'7'-'/ ;~ 141
::I "Y-7:..' ~, =:. Jv7 141 :::1:.-.'~-L.
So
:::I/l-7'7!1- 145 :::~:...'
1- o-;v 138,144,
145,146 :::1/ '~ 155 :::1/r~-l- 132 :::1/.'~/Z 39 ::l / 1~.:::.- 145,155 :J/t.i)"~.i 32 :::l/~7-1- 46 ::I/~.:::. 130, 131, 133>139, 198, 199 :::1/~.:::..::cy;:z 130,131,133 :::1/ ~.:::..:1:/ .A.Al-7 130, 131,133 :::1/ ~;!.-!I 104,182, 184 ::I/~;!. -!I- 104, 145> 1{6, 1{8, 182, 184 ::l/7 :r l//;~ 97
:..//7-1'7 /'A 20_3, 204 :I :...· 7' (..-' 'Y '7 ;z, 144, 145 :::J:..'~-1 1-- 34 :I:/ ;f--;f:.:...' 1- 145 :J~/~ 41 ::I:;..',::;~..:::-7 174 ·\j' 1, 12, 13, 58, 59, 6o, 75, 140, 141,188 -?"-'77- 141 iT-7 Jv 141 iT-~ .::Z 6o, 149, 155, 156, 199, 205 iT-7 -1 / 125, 126 -IT-7 -1 / ; / 125 i_T-7.{/.::t.-.::C.::Z .:1:-126 iT- .-7 /'A 202, 20J, 204 i)"-"E:/ 141 'if'-;t,7 >"l- 105 +j-{ X. ,-;z, 154 -!)" -1 ~1 il 138 28, 59, 157> 187 i_T-{ -?"-{ l- 202, 203, 204 i_T-{ ~ 157. 20J, 204 17-1 J'"t...--;:} 157 --lf-1 Jv 52, 53, s4 -?"-{ >' 137, 156, 157 -IT--(/ ~~-·'f-.-(- 156 iT1/~/ 157 -l_t 7 ~l, ;f-- 185 iJ"7/ v 1-7 ~'l7 145 -!J";b t...-.:X >' 32,111,112 ~~ I:JOH... J:: 169 -!]-'7-I:..::Z 156 iT-7--t.::ZJv'Y 7 156 -!J"'/ ;/:]-57. 97 iJ" 'Y ;iJ I) / 116 -!J"··;>V 109 iT'Y V ;t 109 -!7 ·;> '~Jv 55 i)"J':/7" 89 -?"77-17 103 iJ"-:17-1 203,204 -?" 7'7-{ i_T-{ }< 203, 204 iJ"77/ 41 i_T:fl) 133 133 "7:1 I) ;X /
:z:
r-
-?"~·- 145 i)"~-:::r/-l;l'-l- 145 i)"""?"::l/ 145 iT.:C'7-Jv 46,129 iT 7 ?' 6o, 112, 182 iT7?fffi?Jv 182 iT77.v'Y 1- 115 iT7?'"t...--y r 21,114 iJ"7.:: 43 iT7 1J- 141 iT7 1J -~/ 141 iTO/ 156 i_T /;:};!.- 58 i_T / ;/7 ;:z 57. 141 i_T / l- 31, 32 -?" / )-! 13J, 149 i_T / ~-{ ~1 'f- 1J3, 147> 176 iJ" / J-! '.7 --( 'Y 'f- 176 i_T / l- .:z. 31 i)"/J-'·~-J~- 149 i_T >" 1- 7 145 "T/!~•7o 101 -iJ":. . ,~~?/ 123 $) 59, 6o, 188 ':/7-:t: 104 V7!J- 87 v- (C) 84 -;/- (Z) 84 ~-7 -( (GI) 57 $-'-J:..L. (CM) 190 $.'-_-;(/ 84, 141 :;:,--:;<>.-( / 141 ~:-'-~'/ 110 V-7 -1- (CD) 189
v-1- no -;/-;':/ 34 ~-l~/ 144,145
i)"~!;-$/;t1JS
-;/-"-/ 57 -;)-/-{$?- (DVD) 82 $)-I) 91 V-/ 85,99 ':t:. 101, 144, 145 Y-y;t.)~Y'Y 144 V.:r:7'?.:r:7 103 $.1 .:r:-1 '7 133 ~.:r:-7-.Tl, OR) 190 Y.:r:-:;b- 181 V.:r:-·7.::ZI!:7 172 V.:r:-~ 87,175 -::J.:r:-;;f-'Y:f.A. 119
"7:¥.7/ 35
-::J.:r:-Jv 99
iT~il {8, 49· 138, 139
t·' .I. .A. 'f- ·"1:~:-'Y f- 75
~tt;~:/
35
v-/
j'2,
175
Loanword index Y.~. ·y }-I/~-~/
~.:r:.r~~-~-~
~"\':;,''j'-{!:y
175
f 87
72
::J.:r:.l)- 88
::;·I :;,./ ~ /L/'~ :,_/
}- 142
:..'"\' y.re,y 50 ::7 --\'>"Jv 49,50, s7 :..':1.. -:;( 143 ;;: :1..- :A 154 177
88
:..~:1..-:A:::f-/}- 99
Y.:¥ ?< !J 7. 39, 82
:..-:..'"l''t
~:1..-"'<-1' 64
~7.7'- b.::I/Jf-3!<-/
,'":1.. -_;l/ 99 ~:1.. Y'Y 99 ~:1.. 7-•Y/7. 122 ::J:J..i:.--'.7" 68,122
66 ~:A:J/ 145 ;:.-:A 'T b. 59, 6o, 61, 145, 187, 201
)- 145
:.r"iJ...jf..-7 68,122
~·y 7 49 ~·y/ 141
;:7;-...-¥
:.':1../-?-;v 53 ~" :1..1'<->' 28, 33, 34
43
:'./ .::t :1-/t: 168
C::/~7 139 :.-':::: .:1. ].,--:.-' 5 / '
~.::z.7-!
~- L;f::7 ?
~:1..7-7-lf'~J-'j 52,53
205 b. 174 182
:.-'-1'- 103 Y""-Y"~'-J\
::./ 64,89
Y""-Y-\-/7 '/ 64, 89,90 ~""-/ 69 ~--1'-7'~/~i)j; 69 ~""-.r-~y
69
:..~"I' ;t)' )': 7 38, 172
Y"\'?'JJv?< 172 :..~.,. ;{ s6, 115, 174 ~ "\' 'Y
55, no, 116
:.,"J~.,Y4=- 107
Y"\' ·;; 7 107 /"\' ~~ ij Jl! 118 Y"\'·;;/ 114 ~"\' ~~ ?<- 156, 174 :.'"\' Y
9:-7-'1' /':A 156
~- "\' 'Y #. 48, 49, ;o, 97, 122
:.- "\' }- Jv 118 ~---\'i:.-7- 141
:.-""i:..;v 141 ;:7"";-~.::::.7/ 150 ::7"\':~.::::.-;<:' 150 ~~r~.:::.-:t..-.7~--;7/ 150
53
~:1.. 77 53 ~:1..77-if•y-'j 53
~ :1.. '71VY ::I Y/ 107
:..·s ?:;Jf- }-
178 100, 156 :..·s ;t 178 ::7 s -:..~7 68 /s-o 34 ~s-o>#.- 64 84 /s ::f'Y':f 88 ~ 3 :A?< ::I '1'-f' Y 7- 109 ~!) 59 :J / ;'\ L-~~ 100 ~";v;'<- 125, 126 :.,;vJ\- 140,155 ~"]//'7 83, 175 ~0 ~1 7' 41 / / 95 YY·7:A 154 ~ /' ::I'll, :A 142 ~/-'707"-1'.:1'¥ 142 :./3 -
-:t.: :'./ u
:./::.J/;JP-T-{;('p;:z,-{"' Yij142 ;;J::.J'T-~ 99
~"\'~/ 3135 /--\''7-{;:/J 67
;:.·y Po-L 133
:.,"J~~::.Jlj"
:./::./7-
90
/"\'L. 90 :_... ~ Jl-' ~ Jl.r--;( 99 "/~
"";,/ 52
Y--\'Y/;7-::.·'3 /' 1o6, 130
/--\'/!;1!- }' 57 ~ "\' /' 'Y I 52, 53> 72, 181 ~'"\'/iv{ 64 87 :.-'"\'/!~/ 49,50 :./~>·/~ 87,142
89 p 34> 35 ~j;f 134 ~ /' ;f y '.7 b. 10), 134, 182 ~>~Jv 90 ;:z__.,·- (C) 84 A.'-i- (Z) 84 ;:z,..,·-A,'y 84 :A--{'7. 140 ;:z_-{y 7- 92 7.-1' ~1 7--\-- 75 ~/.::::.3
:A-1 :=:: /' ::~· go, 142 :A '.7 I -'TY 67, 181 :A-'Y 56,110 :A-'Y'T-:A 110 ;z-;~- 59,140,146,148,187 :A-J~-:J/t=:J..-?<- 146 7.-;~-'7-?"'•y }- 59,148 ;:<..-/ 42 7.7J ~~ ~ :1.. 92,147 ;:z,ot--{ 178 7.~- 60,129,141 ;z~-b. 200 ;:z,~;f-;l' 87 A~"\' ·Y 7' 114 :A~--1'7-- 89 :A~--\'/ 157 ;:z,~:;..' 141 :A~/~y'j' 141 :A/;7 I. 7 176 :A·'/I7 176 :A77>7';v 141 A 77 ::.1 7' Jvi •y :7 141 :A/;71)-/ 106
:A/;Jl):J..-149
;;;z 'jl) :1..- J-:'7-1' ;'\-
149
7.7'-}- 144 ;;;z-7-}-~-f 144
7.7'~7 7- 114 ;;;z-:r~- 144 :A-7'.!1/'/~ 72 A ::I 'Y 7-'T-/
97
:A ::I Y 7" 40, 41, 87 A ?<7 178, 184 :A?<-?f:A}-
w
:A7-?i:Ai-l-·~:1..- 148 :A?<- 1- 6o :A7-;<:,y!;J:A 138 :A?<-{ !) -~, ~ :1.. 115 /Z ?< -1 Jv 59, 145 :A?<--1 ;v?'J>J>:..·;v 145 A ?< 7J ::.1 145 :A?< ·;; ¥ 149 7. 7 ~~ )-:' 7. 7 -1'-'r 149 ;;;z ?< ·y 7 73> 115, 194 7.7'T-i:.t 80 A?<;\{> 138, 145 :A? :-
::z.
v:z. v
261
262
Loanword index 7..7-:::z. '7-·r7.. 177 :A.7-:::z.'7-¥ 104 7>:/ ·7 38, 41, 82, 175 :A. 7- ..,· 'Y 7 107, 148 107 7,7--117, 107
::z.::r-4'-
::z.::r-:t;v 143 7.'T-/Jv7 J'/( ~~- 145 ;A.7-ytJ- 83 :A. 7- 'Y ~ 83, 107, 156, 171 :A.7- ~~"f-::fq. Jv 156 :A.7- :::z. -T..,· ;:.t So 'A7-:::z.'7-7"7..177 101,149 A.7->kA.:A.7-Jv 149 :A. J.. 133,145 A J.. 7 130, 131, 133 :A. J..-/155 :A. 1--.Jt• 144 ;A,j-·y7 28 A J.. 'Y / 59, 145, 151, 155, 156 7.. i- ·y 7'? :t ·y ·'J- 176 :A. 1-•;;/?:t'Y
::z.::ry-v::z.
::z.~-tJ-148 :A.~ 1) 'Y 32, 110 ::z. ~ IJ ·.y7 Jv 31 :A.~Y 129 ::Z.7148 ;A.-;1'-1-:::.7 46 7,/-':,-' 101,141 :A.:fl} 'Y'YY- 117 68 ;A.-"{?/ 34 141 A.~-1 l- 39, 100 ;A.#.-7 ;\"?'/ 1o8 .~1.~-·7 .59, 6o, 61, 103, 187
;z...-e.-{y
::z...-e.;v
.A~-'.Y?:r..7···· 103
;;(~/ 48, 49> 140 :A.~>-lJ"- 57 .A:;f.Y~ So ;A."'X'- 1- 155 .A~:A. 115 98 '7.1••... ::/····· 87 ::Zt--7 141 7-.t--7-lJ"-"E:>' 141 .::zt--;v 157 .:r::-Jv7 -1 1- 157 :A.-t'Y 7 115 .A"E:'Y :7" 11•P49 :A.7•yt.J- 117 .A i) - 58, 141 7. 1) -:A. 1-7-1 =f: 141 .A i) 'Y ~~ 176 7. !) b. 140 .Ai) 7 - 106 :z.o-t.J>· 56 ;(o-;z 110,175 :z.o·:; J-;v 118
::z::.:z..n7
:z.
t!-1'77- 45> 46 {!-{::. 40, 42, 44> 87 1:!-ly.- 92 1:!- }< 87 1:!- )l 89 f::-"'7 55 -t!-7 84 -lr-/1.· 7, 84> 149 -t!·7V- 57,108 t!:7v :::z. 7 Jv 144 -tr7v :::z. 7 Jvl\7 ;z.;~. > 1- 144 -lr7VsY140 -i!7l\7 144 f!:::II\/ 144, 145 -i!:JY ¥ 144 t!::I>fr,:.1 F 144 -1:::-lf> )l. 89 t!~7 s ./ 66 -i:!.A "fq.- 175 1:! 'Y 7 .A 108, 146, 156 t!·y 7 ::Z.7 v>' F 146,156 .f!.•y?"Y 52 -lr'Y 1- 28, 58, 59, 142, 187 1:!>:' Jv ;~. y 1- 110 1:!7.:-::::IY 145 -lf~.l't 68 1:!7.:.7 Jv 88,145 -lt.:f-7Jv::IY 1-779- 145 -li:t-7Jv.A. J-~7-1ot- 88 f!l~····· }~ 87
-1! ~·-~ ,. 81 {! t:':'IJ 7 85 {!~I)~ 67 i!7t/ 146 1:!"/Y 142 {! ·~ 140 -!:!::. 52,132 ft ~ 7"····;]; 52, 132 t!~OY?" 140 {!;>(/ 125 {!;>( y I- 39,125 1:!.77-Y 88 .f!.7:::.?b. 88 1:!. 1) - 88 {!;]; 39, 41 -f!;];~ J9, 41 {!;];7 149 -1!;117-lJ"- ~:A. 149 t!o 88 -l:!o so, 203. 204 f!.O::t:. ~ 'Y ~/ S :,.• 203,204 -l:!O•Y ·7 A 139 i!O~ 139 1:!>7. 95 i! Y;- 59, 6o, 61, 157, 202, 203,204 i! Y( - 1- ;]; 28, 49> 133, 178, 187, 188 +!.>'7·21 32 f!. Y 1- Jv"<' >' 88, 175 ~~ '7/1-·7 Jl• 18.5 './ -/( y lj' 149 Y-.{Y?"~VY 149 '/-~ 39 Y-"1 156 '/'-7"7 Y F 156 Y~··:::z.-;v 99
'/ 'Y/ 42
y t:':'::t:.··y 1- 46, 93 Y~.:r. 1- 46,93
'/7 1-7 .{ :A.71}-b. 149 '/7 1-7 1) -b. 149 '/b. I) .:r. 50 v ;v-::1 :r..:::.-y Y 181
-r
-!1-7..
124
11-!JY 82 ?f-1) ...._, 125 .;t-I) y lj' 125 ly.-{ 6o .;t.-( :r.··y J.. 57,114 ?1 -{}:f 46
Loanword index 263 .;/-{-:) IA 1- 181 ($ -1-t 28, 58, 59, 187
($-{~-1 64
($ -1 L. 149> 156 ($-{ L. "7-l::;z 156
'1-·;t/ 114 '1-7 ;z 40. 51. 82
.)r -1-'V 149. 157. 174 ?1-{-'V 136 -!f-1-'V'/::;71.. 1)6 ?1-1-'V.:C> ~ 1)6 ?1-1-'I'Jv 157
"!-"'< 67 "!-"'<"!- s :r 1) 7--1· ---- v .:I.----
58, 141
.;/') /
($:}V- 107,108,182 ($ 7 y - 107, 108, 182 ($:::1/Z 143
9-A.77:t-A.194, 203,204
9-'Y- 82
9->y 7 '7..7!)- 119
67, 89
67 64 '1---l--l\/ 64, 89,129,168 '1-->r -1 ~ Jv 144 7-->r -1' Jv ¥ 144 7--T-1 Jv ~7 -1 ¥ Jv 144 7--T 'Y !- 88 '1--T 'Y l- }]/- L. 88 7--1' / ;z 156, 157 7--T::.-';z:;f--;v 157
($'Y 7 Jv us
?f~y 7./L'::J·---1- 119
"!- .::2. ') 1) 'Y /
?f'Y7/Z7'YJ- 53
~
1- .; ----
82 ($!'\:::1 28, 33. 34. 35. 140, 168, j71.,
170, 173, 178 -!f'r!.>"f- 85 ?17 138
-!f'fl) .::2.- 12 ?17 1J .::2. -v- (WC) 56 ?f:tl) ;I. ---7· -( ----·;t"·--(WTO) 189 ?f:t~J ~/ 172
178
138 :~:f.~ 138
.;/::;,.- 129
"!- s
::I L--- !- So, 132, 151
"!-3
/i:i- 67
7-s /1-Y 64 7-7 =:_;z 82 'f- 1) 172
'j ,._ 46, 47. 71-
9-/
125
($/:}.:::. 149 ?I> ;z 38, 41 "!-- L. 59, 6o, 83. 175 '1-I-'Y 75 7'-I4"-7 125,126 "f- I ::1 j'1. 7--I.A. 181
'.Y7::JY 145 'Y-1-?t- l l l ·-y -{- !" 111
'1-Il.J- 156 7-:r.l.J -:;f---1 156
'Y-fv-;::(' 83 'YI'Y~l.J Y 88
''j -1 'Y
?t-
lll
'Y-1 >' 111 'Y- 58,175 'j-~;z.
175
"!-I 0
43. 88 y -;J 157
· 119
7"-t! 53. 105
7--7
185
7"--t 147 7"-/]/::I-:/ 147 7'-"'?' 28, 52. 99.194
7'4"--IJ'::Z. 108 7-;;:f- ;<, !- 107, 108, 176 7"" 84 7" 7 ;z !- 107, 108 =j·:f''.Y;;:f- 97 7::1-~ 83 7"::1/ }~---- 67,84 7~-l- .50 7'~·-( y 59. 187 7~/'j:] .i 144 7'V9- Jv 8:3.144 7-;; 7- ;V;b .i 7 144 7-:J->r: ·~/.::2. 85 7-:/--1-1:: ;I. 85 7'7..·7 148 7-fr-.JL• 50 7 'Y ;:f. JS, 41. 107, 122 7 'Y -IT::.-' 48, 49. 101, 122 7 'Y /A. '.f--;J 119 71~-~- 133, 134 71~-l-.i/l- 133 71~-~- .i / l- .7. I- 7 133
'Y7- 103,145 'J7-::J Y.;/ 7 7-- 145 'Y ,.--1) 46, 47, 109 'YY-Jv 46,47
39
7-{ ;;('.:::.- 181 T-17::;,.;z 82 7...( t-::.-'"'<' 83,175 7'?'7.. 32 7- 12, 142, 156, 199, 205 7-"7-1:: ;z 199, 205
7b'Y'i<- 83
9-7':1/
?t;v.:c.i -J- 11-< ($ ]/ ::./ !- 149
~ ~~ ---- 82
"!-.:Al::Y-7'::.-' 99 '1-s .r. • 7---1- y:;Jq 65 '1-s ::I 132
.;/.:C-1' 46, 47 lll
·7..( -;; .:r. 1)
7-{-;; ?t Jv 83 7-1 ::Z.7-::.'s ':J:}- 57
7-:,,:f 39 7->-4"-7-.."!. -fv 39 "!- y .- 33. 34
?<"7;.;,
7' -r-- 154.175 7'-(-'r!.-.r..A. (TBS) 190 7-1-'7-1 (DV) 189 7-1-:1-1'7-1- (DVD) 82 7'...(-L. 83.175 7'...(-:.-' 140 7-{-;/ I 1) _:;t:- 82
'Y.::C./7'...(':/.r\}1'7 ~) ::.-' ''j i) -
'Y~/
lll
sz, 82
110, 175
F7 46
77,.- J...->V-1' Jv 143 77 ,.. J...- > v ·"ov:¥'7 148 77:¥'7148
T""7Y
172
7""< 52. 53 7-'C 1J2, 138 7"-'Ci> 138 7."1.-/Z 177
·r:2. ----=;· -1 ----7
1) ----
94
264
Loanword index ·7' ;;t I.').'
)· 94, 181
·r:2 ·;; -tuv F;117 -r· ~:t -;·'
--r •j ~'r
·7' i)
99
99
l· 114
·~ ij •···
156
}·7:/~o 138
:::::...").. '""/'\'"'7
}-7~·{y
~:._,_ -7
136
;<..7 > :;z~·{')' ]../>-::.·}·7>7" 154 j
~) '.'J ~Jt!
Ffl; JS, 144
7-1-· 12, 28, .58, _59, 157, 187, 188 ·r l-'::I 147 71-·:<..:f--1" > 157 7· vI! 59, 6o, 61, 133, 149, 186 7'· ~ s / 59, 6o, 61, 133 7 vI! ;$1 L-·> }· 149 7-v7;t/ 28,59 7-0 6o, 61, 134 7-0 1) :A'L. 6o, 61,134 7->' 157
1--Jl~~}~l 123
7-/~-157
·7->'/7
}:.11.· )· L,Y }· 105 J- v-·1 ~ lJ 7' -1' -
)- v-·1 ~ l) 7° -{ -:~/ ~z 7'
;'~<,;]; 39, 41
; p ·.y :::r 112 }·DL :/ 98 ro>-7> 42
/ ij;t :/ t::::;[;.A 7 85
FY~-<7
)··--·!\/::~-~;•/ 64,84
f1' p ::/ 150 -J-r,'J 139 '}-lj'l,.\ 139 7/ 84,109
F-:I·'~Jl; 1fYfJ- 91
66, 129
)- l:'>y ~7 ;z, (TOP IX) 190
F :;.~ -1'Y 185 )· ;:l .....,'\
181
7 ;/ 7";]; 135 !<7/, ;'\•--· 149,157 ¥7-(:/ 175 )- ;;; ·1
)- ':7 .:z )- 175 F7'Y .y· 140
/•·--· 58,140 ./ -f_J ·;> 1-- 140 /-::J;/ 58
> }·
/ ....."':;:; . -{ -li·----Y s ::-/ 154, 205 j·.{-"J' 154 7-( ,y ..... 8
J-7- iJ-(
146
,/-:/s-lP.>"/99 / - )- 149 / ..... }- -;/ •y 7 149 . / - ! \ '7 57
42
46 82 147
)- ;$1 :::./ 35 )-·y7 123 F 'Y -7 38, 40, 41 )- ':' / 6o, 146 )- •;> ~f7, 143
J-7 ;:?
/ -(P•--·"-t! 52, 53, 99, 1oo
}·Y 39
J-.:;z;\•y 7-1 /'7" 157
139
7,.
)- ::c.;V1'
)- 3'.!1.· 118 !-.A 157
/:1:. 50
; p 112
!-?-;];-:;( 83
F-7Y 69 )- ·7 123 F7?t- 151 F ·7 ?t-.A !- 'Y/ 151
1)
~,~r~
J~~~'Y~··.-~~- 117
l-OY/Jl·-(~~ 98
J<---· r·--· F-'f·•y
;f. 7
)- l..--..-'{ 147
}- t,~-:./ >" ~7 147 )- v-;'..' / 7..--{-r;- 147
}· r;:-7- 110 ;<. 'J ;]; 110
111
7.-:r.L, 178 3/'. ij]/ 7 )- 139 t. 'Y )- 132, 146 3/'. 'Y l-7/::t 146 t. 'Y )- '7 -7 104
L. 201
)--(;I:; 64
J--7-iJ
;'r..----·7'-{ 7',-;.z ~---·;J_J•·-- 148
201, 201
204
l· -r v'Y J· -·'\·---j'~·---· 100 )- :::C./7 -1- 111 )- -,( ·;> ~7 (TOEIC) 189
··:__,_ -."3./l--7 172 98
."-' fj'- 50 7,. -7·-{ ,. 148
28, 31 34> 35, 168
F 7 104, 157, 176 F7 :r.:;;:)y 157 F -( 'Y 38, 41, 68
.:::.::2. -3-•7 6o, 172 .:::.::tll··>-"~)[;~7
51
7'' l) '" !j ---,;J;.:z 156
v :/
136
156 I.-( ,:;z 178
} - / ( :/. - " ' : .1
7?t ;l;
J2,
77 .:I. 7 /[;
./ .J.f:~/ ~;]; 7. ~· 85
/' Jv"' 88
28, 46, 47, 91 />".58, 140 / Y-l!7Vs / 140 ; \ - 128 i~- 104,134 J',-.y· 88
33
-::I-F 155
117
.:r.- )- 55
7·y-;..·y.j' 117 7'Y ·-y 110
~~-:./")._
7"1'77-P'J 7t: ~) 67
~~-/>n- F 157 l':-.-:.z.::I / J-- p-;]; 146
84
7>;{!:;..-;z, 58 -;J--:.n~-
6o --:.-:A' 57, 89, 110, 154· 200 .::...... l· (NEET) 189 :::.;b F 146 .::._-;t_ 135 .:::. 'Y ;:I:J F 146 :::::.. 'Y ':r /[; 40, 122, 14.6 .:::. -;r 'T JldJ F ~ 'J 1" 146 :::::..::1.- 1,)6 .:::. ;:. -7, 57, 8o, 100, 149 :::..::J..-7A-r;z.:Sr- 149 }· IJ / 140
--:..-'>
157
1'1-')7 ;J; 146 ,l'~-'))-.lt'::I I
i
> ~::J. -!<-
\-7" -(- 72, 156 ~ :r 7';]; 169
~f&i 't ;fl 169 !'\'"" }-
141
~~- )- 149 }'~-]--
.;, /, J:..
!'\- )- 7
)]/
149
141
;-,-] 156 !'~--;/';]; 82 }'~-"' 57> 133, 134 !- 133
i ~---·v
·*' :/
146
Loanword lndex 265 1 , .... 'E:·-T -{ -tf·····;'./ s
/'
'"''~ f', 87, 108
:.w3,204 ,, .... '!::·.:::::. , ... 1)8
/'\·.... 1.--· .... :~I~>Y ,~ ....
)~':/:/
100
;'\]..,
97, 185 116
1 ,.;, ~~- 58, n6, 141
rq/- r
;·,-.y 1::::-::r.::./ ¥ 141
;\[/-;J~-)[; 97,100
!\·;> 7 T- 117 /~:-r ;1; 31
l'~
]/.;,J _,.
y 142
r~- V/·r-'li 131
;\·7-r..--- 31
'' r..--5 -1 >7- 142 r.o- 146,157
1 ,.,· }40, 144
;'\7· \..•> 31, J2, 31111,170 ;'\'f·>'vY 31
J'\]::l•----r;r·---7 157
]..,·:/ 134
/'~ ...( 64, 75> 84 1 ,_, • .<J·-r J.~ j }-
IJ- 153
1'\r:J-t:-.:-~::.llj' 146
~~ J-. ;fJ- 144> 145
l\T.17Y 40
! {.{ ;.t ij ::./ 85, 183
,q.. p ..·-;J;
;'\-{~'Y7" 69
J~}-P-A1 ;.b-144
;~n·T"--r- 139 1\T:lt-=- 146 r;o.~ 139
!'\-{7---r-Y 140 ;'{-{ }- 28, 52, 135 ;'\-{ ;\-{ 58 IH/ 155 i"..-{';!'7-b 150 IH:fl.-,-:7.- 150 i"..-{-;J'j,..--:)1-77-J..>. 150 !V( "f,... 103 ;'\-{I) Y;iJJL• 92 !'\7 .7. 55, 149, 157 1~'7 :<Jt..... 149 !~"l ~ ....,~7 149 1 ~·'J 7- 100 r~::c. 1) ·V 91 1 \;i:J y 7, .so, 101 /'~ :7 123
;\7'-:'./s /50 ;'\fr'Y 55, no, 171 J\'r ·y }- 110 ;{fr-y }- 1.21 1'\~/;j 109 /\::./:/:1. 109 ! '\:/' ·•y :../,:!. 109 r;::./7 -1 '>' :'7 1) - ':! 147 !~:'.-'-\:'""?' 141 i"..::Z 115,132 !~A7'7 32 1'\A::JY 146 !~A5 43 ;~j :l j 146, 198,199 l':9- 42,117 1 \7--f' ]... >7 1) '\7 Y 181 /'~';I ~7 149, 201 1 \ •;> ~r ?.)·7 .{ ;z, 201 r~-:~7=f-I:''EY 90 ~'7 149 1\'Y ~ 1 ~ .:2 -·::,'J.y 7 149 ;{-.y 7: ..... 104, 116, 174 J~y'f·
67
/'' 'Y 'f-;'\ '>'
:'7 119
r\ ·;r'T .-( Y J}' 157
;~.::::.71i>
144
138
/;.:::::. 'Y 7 138 1 ~.:::::.. 'Y ~7 -t>
138 St.J- 87 ''·*.b.-'·./ 44 ~~~~
ss
1'./:f.J'- A J.. 7-1'* 91,144 ''~'t.J 1- 55, 124, 125
1~7 149 1~7' lo6
1~7 123 1 \/
/'\0'7 157 28, 31, 32, 65 1~> ·~bY 65 1'\j;;IJ'- 144
r; Y
7<-t- 32, 33
!\/'.:f:J'f------7 125 ;'\'j-7 44 l'•>"~lJ]; 67,84
/ ~:fT -{ ;.z:'-t:- 32 / ~'"/ ~) ';I 7 104, 203, 204
; '>'r'f- 124, 125
!~';fl) ·;>~7)'7:r..7-7~· 104
;,y'r;J;'f--7 124
1~-;/'1) 'Y
7 -( >'Jfj\17.~~ ::./
J-. 203,204 ;'\-;!';]; 8.2 !~7"]::1 106
/'\""?'"" 90 /\""?'-;;z. 87 1 \}:,,
JS, 41, 87
/\b.7:7- 90 /\'C.:&> 138,145 144
;\7 J:"T -1".
~~-7.:C:T-t-7-1"' ~/!1 144 1~7 ·':17 --{ 181 ~~-7 A
125 ;\7A }- 125 /'\7 A:!.
Y }- 144
;\7 f ;]; 144
;':77--1 t.J
45· 46,129 1 ~ 1) 6;) 87> 178 /~~ l.J
;'\ i)
7 157) 205 '? 7 I) ..... 157, 205
;'\l) 7/';J;:.J:;I.:r>·~t- 145 /'\ i) , ... 146 ~~ l) -lj' 147 ;·,i)-#.•y 5'- 146 /\ 1) t.J / 68 1 ': i) ::1 :./ 14.5 ,I) #.:7.146 ;':;J;"f·:: 7 121 ;\t~:r. 185
r.>'rf~-7 125
;{Y::J 31 ;'\/:.17 68 !'\Y::h1 J.G 88, 109 ;~>·~t- 89 /'\JA }· 144,145 !~>"'J 107,144 i\Y·r-{YVY 91 J'\>'f 144 i\Y¥ 142 J:-1' ¥ 68 ;'\:;.1 }''"? /' 142
;'\/' r
!'./' P.lL' 154 ''j7';v7 105 /'./'""?' 176 ;,y""?'- 90 ~-7-Jv (PR) 189 ~7if-7''> 42 ~7.-.z. 104 ~7 / 174 ~77 122 \::-55 ~-.... 12, 13, 86 t:::-::r. -9'- (pH) 189 t~-."L.J.>-.1/:f 1) ;•.- (BMW) 12 ~--:r-
(PK) 190
t~-.'A' 110 ~-;z, 175
1::: _ .'T -{- ."L.- (PTA) 1.:: ----7- 7-. 32
189
266 Loanword index
t:- !-' '-" 34> 35 2:-7 ::Z.'T-4'- 124 ~-"' 50, u9, 155 ~-{ 1) 8 169
!:::/;(- 83 1:::/ f. 29,149 ~=::,.~ r ?- 83 !:::/~/ 125
~-I)
7 • ://3<"!'\ 43,108 7 7 ~ ·7 ,7., 12, ;8, 59, 108, 188 77~Y·s >1 156 77~$./s /"'VliA 156
yJf57
t:- J]; 28, 38, 41, 42, 140, 170, 171
t:-'7 :-.' (Bl) 189,190 ~'if' 43 2:~-i'
81
t:~:=r
:-.' 200 t: ;:;z ·7' ''J f. 150 2:/<."f-;:f- 55,123, U4 ~.::Z'T!)- 99 ~A f.;]; 41, 153, 171 2:7.~ 32 2: A"'< Jv7 172 t:?t ~ / 175 t: ''J ij' 133 1:: 'Y ·7~ ~ J;J 133 ~'Y7'--\'- 117 ~,y···y
7 43 ~ ''J f. 146, 178 ~ ''J f./~ 1..--- f 146 ~ 'Y !~ 1-· 146 t::.'Y/ 178
!::.::. 130 !::.::. - )]; ¥, 49> 85, 150 t:.::.- Jl•-J- -1' p > 150 2:.::.0 130,139 t:.::.py 150 2:~/l'; 67 2:77-.::f 49, 124 t:..:V 42 t:..:V.tf-7> 42 ~""r / 174 ~.:1--;::( 86,177 t:;;!. -!::: -r· 109 ~.:1--"'< 93 ~:z. ·y"f- 52,122 ~s >"""rY 66, 84,90 V:G? 170 t: ;]; 58, 59, 132 2:;];7-{ >7 59, 132, 175 !::: 49, 95, 103 t:p- ).:' 34> 35 ~py·;:f- 46 ~/--(/ 64 ~ / 7 91, 139, 156 t::/7t.r~ 139 ~>'7-iJ"P>' 156 ~>7 L- r 156
c
v
7 7"7!J -::t.' 138 77"7:f> 138 7 7
~ ::I ::./
145 77~ 145 77~ 1)-:::J/l::::z.-?t- 145 7 7 JV'Y 89 77>' 16,56, 86 77/ 86,87 77>'-7 191 7 7 Y--'T-$.1 s ./ 101, 191 "7-1' (V) 86 7-{7/-t 173 7-1 I.,:<, 7 72 "7_.{"f- "f 7-Jl• (VTR) 189 7 -;' Jl-·b 86, 107, 175 7 _.,, Jl•b 86, 175 7 ···( y7J~1.G?.!l: 185 7,.,; >7>-' 68 7-y';:f-/ 93> 109 7'-7/ 94
u-
r
7rr-7 104 7:-t-;];}- 86 7;t-;];}- 86 7:.t7J·'J'f--\' 88 7:-tJl·'T 43
7<1"'777 l l l /-lf- So, 129
/"-tTY 66 /~7 7 149
7·::;-:/ .:r:.)Ti
118
/"~1-:)j lj'
125 /"ff-/7 125 7~ f. :;f.-;]; 146 /~l~-3<-J]J 85 -t:::r-1 >1 7" 125 7"1 1 Jv'Y s9 /"7I..::.~ 89
7 .:>.-~s Y 181 7 :1.-::t.' 86,177
77-1 144, 151 7'7..-( :- 7 203, 204 /"7/, 7. 141 7'7-{A;t7141 77..-r r 91
77--( ;~..-( 12 7'7..-( ;<;-::-·- 57 /7'7.:<, 56
777 135 7'77" 133 777Y'3 >' 135 /7$.1 55,109 7'7.7.. 157 77.A::J 3134 7'7 .A 'f-~ 7 147 7'7.A'f-~7 7'E:--'TJ1· 147
~-~'.!no
7'7.A ¥7-1' >-\- 157
7-f 87
7'79'---T 40 /7~ '7 ;8, 6o
/-L. 140 /-:J-7> 125 /-:J-7>7125 7'-.!l• 107 -;1-p-.::.;:z. 91 7 ::r:.-1 ,'2, 178 7::r:.-r7?J-- 72 7::r:..A 135 7 ::r:..A"f--{ ;~;J; 135 7 I . ~.::..7.. }- 155,175
7::r:.I)·Y7.A 122 7 ::r:.Y:--·Y-7 181 7IYA 86 7:S:./7. 86 7:-t-iJ.A 72
77 'Y ::..-· :t
107
/7~:-/ .:L 109
7'7~ 7'7~
l--7:-t-b 86,175 l--Jt-L 135,175 77-'f-J]; 41,86 7'7'E- 147 77>'7132 /7>7~ }- 152 /7/:::J 35 77/.A 48,140 /7> F~>' f. 149 77/;;f.Jl• 39, 41,86 7 1) - 149· 157, 205 7 I)--!]-;(;::( 157
Loanword lndex 267 7 7 7 7 7 7
i) ·----~·-·· 149, 150 l) .... .;i'-{ .,, ;]; 157 i) ...... 7 ..{ ~;,: ...... 149
JV 150 "fi) -t;1/ 150
7" -y"f
41,107 7 i) ~~ }- 88 7~1) 7-/' 106 7'1) :t •:; -t Jv 122
7"!) /
7"1:17 .I~/3"Til' 6o :lo·"'\7 13.5 /"U;J~-.1' 137
7•r:n, !J: lJ 7- 1- 6o 70:-../ 1-- 148, 157 7r:J ::..~ 1-757 A 157
~/:/)];
~.>-7"-{
174
·7 157
~-A}·
.i~- 83
142 55
-'\-J--~i'.I/ 183
•'\----· )· ----~>
183
-~-r~- 147, 149
(pH) 188, 189 .r'\::/ ~- :i'fL· 142
·'\·--'!' ~A
~7-.
l- 139
r-
7t/"tt:/ r- 59 71..-·> r 146,156
:to
6o, 61, 136, 148
7't"J-..::.;/-:;f:·./
10<)
7o 'f
115 :tn~"7L 6o
7''o A ·r r
::1. ·---- }·
~to -t -;rtJ-
7
141
--~'f-;1:} 46
6o
:/o-y 7 115 ::to-7" ;•. -+J- 177
;~ -'!- F :r. 'Y ~" 142 ;f·----7' L~ 31 ;j;--=?':;.; .51 ;f- f 125,144 ;j\42
:,' Jj'
;f:--r A
154
;t-L. 86,135,147,198,199 ;t-L.-"\-9 147,154 117
r rr
~77> 85
;t-L.~;lJ;~- 199
~-7>
r
;JZ-IJ /
f
182 185
::t-Jl-· 157 ;t;-;v 97,125,149,157 ::Y-Jt-·~-> 149 ;t<:-;V.i~-1 / ]-~::-' 149 -~~:Jrv ::.~ 1f 176
·'\-7·>' 63
;JZ 7-tr·----
-~-=-
t. 89 •'\.::. ·'f 174 -'\.?f.. 'f-7 85
;f~'Y/7" 176
~P.:- 1,)6
#.7•'\)]/ 69 ;J~ .:r"C > 144 #..A. l- (POST) .56
~1:~:-+J-·7 J]i 143
144
;JZ------1 55, 147, 156 ~~-~ 1 104 )~·----.A' 157
r- ;];
39
~§yf--y
,""\··y/l·~~-:, 1 124
134
88
;j;-{ 'Y A}V 93, 118 ~ -{ /' 1- 59, 6o, 61, 149, 157 ~-{ :;/ }· 'f 'Y -' 157 ;f<:•) IJ ';/ Jf 185 .i~t) It• 182, 185 ;fz.I.-J];t):>J•y7~::-'~' 93
~-., 'J 7--{
::.1
69 / Jj' ;8, 153
-"\-{ :,' }- 175 '·'· 59> 109, 147> 168, 178, 185, 187, 188 ~-9;!. 107 ~.-;z, 146 ~-.:Z.T•y/ 146
"'''-'Y )· 41 .,-z -y }- 51, 115 ~.·:; 51, 73,115 ~-y 156 ~.·:; -7? ::..~ 22
/1..-·~::.~7-Vs
r-
28, 46,47
~7-T-{ J: l l l
7'l-·- /
7t/{!:;...t 134
~j
105
--~-19 185
--\-----:z;JZ--·--}v
145
>
~/ 149> 157 ~->ot 41> 107
·"Z•y '!-> 135
:tv::I> 145 7'v-::''*-7 87,109
,.,.,JV~ 109
.~,7
~·v-:·_,~ 175
r- 7 _:;z r- 175 :tvot--v A r- 145 7'vot--v A r- ::I~'7 1J- r-
~Jt.·?t'> 101 ~,;]11~- 199 -'\ilJ-'~, 'Y '!- > 135 --~;VZ 'Y f· 35 --~;]; 1)
103
91
156
~,;];_:;z
"'1-'A HJ-1 ;O
~'"'\.7 /[.1'./
106
93
> ij'.i~ I)
74-iflv 172
~7 157
s,
~Jl.•V
7'0 :-/ }-- :r.~. ~7 148
...-,7 t::'/
v
7t..-·-A'
~-;j;J;-:/
148
~7 146
12 126, 129 104 /l) / 7- 104 7 ;1/ 1,141 7'/L------ 58, 6o, 145, 156 7';Jr--· ;z, 99 7Jt·----·'j 110 7;1r---'f'·---- 156 7';11-"<')::.-'7/ 145 /;];::I:f' 67, 84 /'J[;-7-:: :L 105 7';];-7 '7 85 /'J[;-y 7 109 711' !-' cY ~'7 115 -:/}1.! )-' 'Y 7 115 7117 "< / 145 :1L/- 55, 149, 156 /l-·-£-1 156 /'v-"i=- 107,157
> -7
-~7 135
::to7.-r-;l!
l) ----7:/ _:;z 149, 150 i) ·····7 :;_/A 7 ;];;'\-{ ~----- 150 1)-7/A·;;H-!J- 149 7' i) :Jr 41> 107 7'1) +f- f 150 7' i) f- f 7' ~7-t
'1'1)
:fDI'~j:j':;i ?f 60
7"1:17 148
~ r.>:- 7,
l- 'Y "j' 1,)6
108
;~f'Y }· 144 ;~7-y 1--t->::Z~- 144
.....,.;]~/ 124 .~.~ 103
;j~ -!7-9 109 #.-!7-9 ;l. 109
''v'l"t::'> 174
;~!J:> 28, 34,35
268
Loanword index #.7'.;,/lj 125 *'Y~- So
*'Y·7 41
-
* *
'Y J- f 'Y 3' 97 *'Y/" 41 #.·y/".:z. 143 ;f-'T-{ - 145
-
*7-ft.- 55, 59, 6o, 61,129,150, 156, 187 ~~.:::. 91 *L.~ 147 *7•··· 87 ~7 :.'7-·1· 7 181,194 #. I) $-'-· 130 ,f._ I) ;z, 105 ~;vv ::c. ~.:f 46, 47, s5 ~JvV7 46, 47,107 #;v~'l :ii.1v 34 #.;t.- 1- t.J;v 34.35 #..1v/ 102 if.Jv/ 1'77 -r- 102 .~.Jl/7 .'7. 7'.111 7 85 ;f;;t.-OC/ 53> 107 ~t·- 97 #.o~--1! 91 f£ '6 ....... :itt: 111, 169 '7 -1 l- u, 9:P56 *'7-1 l-V'i''Y 124 '7 . .-( l- 7- 12, 156 157 */:::I/ 64 #./.A 38,39 ;!- / ,-:;(' 38, 39 #./'7--i''Y-7 67,84 ;!- / f 35, 38, 39, 40, 41 #./7' 39> 171 #./•A\ 53 ~/-"\-{ 68
*'
*
*/
"<.."-7 14, 104> 108 "<.."--;)-\'" ~/ 64> 89, 90, 168
"?'-.A J-1) 1::. !- 109 "?'-~-~-7 63,64 "?'-.:C-;.r !- 38
"?'...'( 1,58,60,140,198,199 "?'-1 '7 57. 132, 156 "?'...-(70 140 X'-{ '70?" 156 "?'...'( ?o-:7-va ./ 156 X'-{ '70*:-' 57 "?'..{-;f-L, 140 1, 198, 199 "?'...'( "?'7/7-/ 145 "?'jj-;)y 85 "?';bO.:::. 157 "?'t.;O.:::.?.:r...A::It::./ 157 "?'-7 F 133 "?'7 ~-T1vF 133 "?'-'f;?f...V?b 40 "?'-:f;?f... 140 "?' 'Y '7 $.'I..{ '7 133 "?'';l :;t. 1 87 "?';v7.A 140 "?'Jv::::I • :;f-o 183 "?';v~-:::z. 107 "?'Jv7· 140 "<'Jt··r ;J, 32, 170
*. . .
v
"?' JvJ. o 34 "?'0>" 101 "?' / 69, 75> 141, J42, 157 "?' >"7 32. 33 "?'/1~'7- 143 :::-(-IT 32.169 ~-·lj 32 ~ ...'( 7 33> 34> 129, 184 ~-IT 32,33 ~V/ 149 ~.A 8, 133. 145 :::..A::::I/ 145 ~.A::::I:-./7-.A !- 145 :::..:z.::St- 147 ~.A::It-F--J-'Y 147 ~.A7-7 133 ~.AF 147 ~.A'"'<'Y
~ .:1. -9 ~' 7 149 ~.:~.::.'""/ 67 ~ I) 27, 28, 49> 1J2., 186 ~ 1) :/.-I- Jv 27, 28, 49, 132. 178, 186, 187, 188 ~ .fl,·'7 141, 151 :::. ll1'7 $-'I. -.:f 87, 175 ~ Jv7 ~-.:f 87, 175 J- 151 ~Jv'77:s ::::1 ~Jv7..(-:::z. 48,50,98 ~ p-;...'.:I: ~'Y
v-
~ ::./~./
55
:::.Y.A 55 b-7..( - 154 .b.-~ 156 L. 'Y7- 73 .b.~~ 91 L./1~..-( 68 :/.7 ~ 147 :;I.-{/ 185 :1--t.J- 6o j. -'7 149 :l-77•y-:t 149 i- - 1-- Jv 28, 49, 6o, 61, 122, 178, 187, 188 i- -111 14,147 .:>( ...·,t.-7 n,,;z 147
Loanword index 269 .)(-:.' 185 ;>(->'7-1> 75
.)( 7 .A. }- (MEXT) 189, 190 ;>(~.;.-
141
.)(~1'-.A./--·/
141
)<.A, 39> 122
j. -tr-r 48, so, 194 j. }l ;;j; 133 j. :7 ;;j; l) 'Y 7 133 ;>( :7 ;f- l) ~' 7:.--·y PP-b. 133 ;>( 'Y ;b 94 ;>( •y-1! 53 ;>( 'Y -tr--;:7 117 ;>( l"-"\--::l.::r.7 94
.:>(.:C 1) - 148 .)(.:C 1) -.A.7-( 'Y 7 148 .:X'f:'-1> 138 .)( 1) 'T> 124 ;>( 1) 'Y }- 200 j. 1) "'V .A. J4, 35 ;>( /Lr"'\ Y 52, 87, 141 j.;V.,Y';.'-•y 7 141 .)( Y7- 55 j.:,--9'-;f--j[; 55 ;>(Y'Y 62., 64 ;>(>7 135 ;>( /7~7>-:.<, 135
.:C7l-·
111
'r.-:.z.
102 .:c-:7- 150
.:c-:7-7 r~- }- 150 .:c- F 157 .:c-.::. y lj' 146 "C-Jv.A. 102 .:ct.J 147 .:C.A. (MOS) 190 .:C.A.7"J'-y 85 .:C.A. 7 '7 68, 85 .:c:z. :I 68
-;v
68 .:c:z.::~ ~ :1. .:C?i'Y 147 .:c-71'> t.J- Jv 147 .:C?f>;f--..{ 147 .:C'Y :t 114 ~Jv 59,147,187 .:c. }- 150 .:Cl"\/'"'?" p 172 ·:C P.. J9, 40, 136 'f:-:;J~ 147
.:c;v t::.-*- 39, 40, 86, 1)6, 170 .:c;v.:c '-" r- 38, 40
7/'>t-7;v 156
So .:c:..7. :7- 144
71.>.'.7-;v 142 7b.3f:.. 124
-'E;::.':f--
-'E;::.' r- iJ ;t-;v
68
-v- F 93,125 125 -v-Jv P 123, 125 ~----j[.;
"'Vb.7-.;. 64 "'V>:f> 68 :L- 1,60,99,134 ::;..- ;b l) 13:),134 :L-;b 1) /·7 .A. 133 ::;..-}- ~7 150 :L-7 :.r- (UFO) 188,189 .:::. 'Y 'T 67, 93, 98 :L }- VI::: }- 98 =-7 89 :L*:Z. ::I (UNESCO) 189 ;1, !) ..{ ;tJ 184 :s-lf';v 102 :s-.:r·;v 93 :s-::>7 b. 168 ::3 'Y7.ll' 150 .3 'Y }- 150 .3'Y 1--*7-;Tt- 150
r-
1
7-;>(y 64,168 71 ;\ 48,154 7-·( :7- 149> 176 7-1'7·- 64 7-1 }- 157 7-1:1 157 7-17 :z.?l-1 Jv 57 7-1/'Y-( 'Y 1::: 72 7..{/N:J.A. 157 7-1> 100 7-lf-.::..,. 91 7-!f.::.-'\' 91 7:::-~;t 8, So, 145,146,173 7-::>;t::~> t-o-Jv 145 7~::1/ 145 7~-\' 33>34 7:7 }- ry..{ ;L 87 7 'Y :\'-- 58, 142 7:Y::f--l!:f>142 7'.' 1 ~~ JS, 41 77.:.1)- 82 7;;- 144 7:1 115,156 77.* 156
71.>. 14,142
7
1) - 129 7>"142 7>':t->?" 91 7Y?"·..·>' 68 7Y~.::r. 1)- 50 7>¥ 4], 156 7>1"-tr;J.. 41 7>l:::':t- 123,124 7 > / 41 l) -lj' ;6, 147 1)-.A' 88 IJ-Y'7:fJ1• 58
1)--"\ 52 I) ;b- ll7
lJ7;v-t- _; 1Ht..{ 7
Jv
138, 194
1) .~7 200 1) ;z, }- 12 1)
.~ 1--7 133
7. 1--7 7 7 7-.::2.7 1)
:.-· lj' 133 1) .A.;f-> 68 1) ;(b. 107 lJ .A. ~J / 124 I) '.1- }- 144 1) '.!··- }· '7!'\• .... 144 I) ']::y,r; 144 1) 7-'.J)., 40 1J 'Y ;Tt- 49, 122 I) 77 $.-'- 202, 203, 204 lJ .::.:1. -7 Jv 168 1)
r-
'" e 1J 133, 199 1J "' e lJ -=r-v s > 11 1::: 7-/$.-' :I. ?r ..{:;_1
133 109
1):1142 l) .;f-:;.t 129 1) .:c- }- 144 l) .:c :I :;.t 144, 145 1' .,.fj'.::. 93 l) :I.-~:I. 99 l) :I.-1-- 100 IJ:I.-'"'?"'7-142 1):I.'Y7
52
I) :I. 'Y 7-IT'Y 7 52,129
::r-Jt 91 l) :;.t;b:;.t 172 1) y lj' 12, 13, 142 l) /::Jj[;y 172
1)
1) y '~ 39> 40, 51 Jlt-{7-( ~) ~'l / 121
270
Loanword index ;l!..{-": 66 ;1;-·y 57, no
;li-b 88 ;l!-;v 91 /Li'Y ·7 156, 157 .!l''Y 7 :::Z 97 /[;~Jl-'·9"~::.~ .:2
]/:::Z f.7Y }- 126 1.--···y 7,Y 106, 119 t--HY T /1/ 4J., 122
p";"':;../ 171
vr..{-141
D 1) :J:.-' 145 pyl;J" 140 D>f=.7 101 DY!-'Y 67,74 DYIJ- 101
D
!.-7..{77-.:::Zl- 141 49,
1-·J-..Jl.- }- 39 V;~- 157 v~:::z.-148
105,122
)\1:.-'·-,:/ 53
v7
]_,· ..{ 7 ':7 }- 182
v/7
v-1'/
v---\J\1 141
185
145
'7-7 157
40,42
'7- }' 144 '7-f:fp-(!y
~/t:7;;7 141
v7'::z.-
181 ]_,·-.A 59> 100 v-=.:/ 172
Vt"::- .:l.- P 124 VCY 89,147 Vt:c Y :::Z/b 'Y /.2 147, 176
v-Y
l/E-:.-'A·7 7~~/ vY9 101
v->,-~ 39
-?t
109
]_,·..{ :;.;::;-)- 35
101
1)
D 1)-)t ::J:.-'/1.-"Y 1;7 :::Z 145
176
-17-144 '7-/D 144-145
'7-1/--1'"1 55> 124 '7 ..{.A ~-;1, 85 '7 -{ f 156 '7 ..{ }' ~ 3 156 '7 -{ -'f 103
vl;Jl) .I.-~e :;.; 194
]/ y :7: 28, 39 vY t-7::.1 51, 52, 98
V'f-1 185
D-.A 55,124
'7-iY 92
v:I-?1-
P-7~
'7 ;b 34, 35 '7 1;7 7- ::./ 51, 52, 88 '777- 105 '7/Y }-- >" 6o, 91 '7 ·y 7 Jv ns
v.:f'Y:::z
125
147 1.--'::I- ~ 57, 142 V+f- 87 1---:::.~ 132 vY.:z.?t- 132 t.-.1:./~.:C.>'
}-
}-
124
D- f 157, 203, 204
D-f:t"/"tf1)Y:7u D-f/7-1'V>"7 203,204
D-Y 101,154
'7=-.:::Z 135
203,204 1---:::.~--1'- 57
D:7/,Y 195 D-7 156
'7 )]Jy'--\' '7 6;
V::Z/b 147 v /. 'T Jl' 150
D-7-~:s
'7 Y.:c;v-7'-1' --7(ILDK) 189, 190 '7Y?tY 64, 84, 92,97
202,
v::z }- ., 7 y }- 126 v7. t--7Y 126
v :::z t--7 ::./ }-
v:::z v:::z
126 t-:t.-Tiv 150 t-7> 50,126
Y 156
D-1! 1) ;t 32, 33 D:::Z-7 46
D:::Zf.D;¥-rf..{·Y 7- 123 D'Y ·7 91,141 D'Y ;;:f. 109
'7/
1, 58
-'f.A"f 172,176 of .:z."f- 172, 176
Subject index
A abbreviations, Roman alphabet 186-188 Abe Shinzo 191 Academiefranraise 206 accent, Japanese 3.21, 74, 126--130, 134---135 accommodation see adaptation accretion, morphological 142 acronyms, Roman alphabet 61, 1J0-1J1, 159. 186-191, 199 adaptation J, 10-13.21, 29, 41 52, 55. 62-65, i'l· 75-126, 1Jl, 1)6, 141-1.43 adjectives 1-2, 21, 31, 58, 6o, 137.139 adpositions 1 adverbs 2, 21, 58, 137, 139 advertising 1, 13. 19, 166, 168; in magazines n, 20-21, 27,58-59,186,188;in newspapers 17, 20, 159, 186, 200; on television 11, 17 affixes 1---2, 21, 59, 139---141, 15_5; see also prefixes; suffixes affricates 73. 81---83. 88---89, 1.0<)-110, 113~114, 1.19~120 Agency for Cultural Affitirs 163, 193, 195, 200-201 Aguranabe 25 Ainu 2, 7, 23> 36, 66---67, 166 Aleutian islands 45 allomorphy 150 allophones 71, 82, no amelioration 154 AI~iro
29
aphaeresis see fore-dipping apocope see back-clipping approximants 73-74, 81-82, 91··-94
Arabicloanwords 43; numerals 159 Asahi Shinbtm 15, 17, 27, 160, 171, 18J, 185, 193 aspiration 83 assimilation see replication, antidpa.tory Austro-Hungary 43 B
Baba Sadayoshi 45 back--clipping 130---137, 148, 191 bamboo English 57 basic vocabulary see core vocabulary Batavian Republic 48 Beach Boys, T1u: 126 bilingualism 2-3 bimoraic articulations 75···-?6, 86; spellings 175-176, 179-182 Black Ships 42 Bonaparte, Napoleon 48 Bonin Islands 57, 77 borrowing see loans Brazil 76 Britain see UK British Commonwealth .56 broadening, semantic 31, 140,154 Buddhism see Zen Bunka Shingikai see Culture Advisory Committee Bunkacho see Agency for Cultural Affairs Burma 44
c calques, Sino-Japanese 6, 10, 42, 44. 62, 169-170 Cantonese loans 63-64 Catherine II, Empress of Russia 45
Cath()lic church 21, :.1.9-32, 169; loans 29-33, 43; materials 30 Chamberlain, Basil Hall 54 Chang'an 62 change, semantic 22, 55, 140, 143, 154-155; see also broadening, semantic; narrowing, semantic China 6, 23, JO, 33. 36, 44, 55-56, 61-65, 76, 160, 166 Chinese characters 6, 8, 41, 61-6-t, 153, 159-160, 166-17J, 185-188, 197; cuhure 6, 8, 61; loans esp. 61-65, 84> 89-90, 166---168 (see ~0 DONOR WORD ll•"DEX; h~-uontenlci;
wameihyoki); Revolution 61; topolects 2, 23, 62-63,166 choon 169 choonpu 162-163,171,174-185 Christianity, suppression/ prohibition of 30---31, 170; see also Catholic; Dominicans; Frandscans; Jesus, Society of; missionaries; Orthodox Church; Protestant Church civil service, Japanese 206 clipping see back-clipping; compound clipping; fore-clipping; full clippings; half clippings; mid-clipping; mora-clippings clusters, consonant 21, 75, 106, 119, 124; media 18---21 compensatory lengthening 121 compound dipping 22, 130-133, 143-148; reduction 22,
272
Subject index 130---133, 143-··150; see also ellipsis; portmanteaus compounds, assembled 11, 14}··144> 14?··-148, 151···152, 155; hybrid 33, 38, 42, 52, 6o, 63, 74, 124-125,146, 149-152, 190; imported 143-144> 147-148, 151-152; made-in-Japan 143; semantically remodelled 11, 22, 155···157, 188,199 Confucianism 6 conjunctions 1-2, ;8, 142 copula 139 core vocabulary 1-2, 5 core-periphery model 4, 9 Culture Advisory Committee 180
D Daigaku Toko 51 Daikokuya Kodayft 45 daimyo 30, 33 dakuten 161, 165 Dejirna 33, 36, 38, 47-48 deletion 21, 55, 71, 81, 123-12.6, 131 dialects, English 102; German 105; Japanese 12.6; Kyflshft 31, 38; peripheral8z; rural 82; Spanish 81 dictionary traditions 62, 76-80, 83-85, 89-93, 97-102, 105, 108, 115, 178-179 digraphs 148, I6y-166, 174, 177, 181---184 diphthongs Ioo---104, 113 Doeff; Hendrik 48 Dominicans 29 doublets 34-35, 42, 50, 65-66, 82-88, 101, 104, 107-110, 118, 123, 146 Dutch East India Company 36, 47---48 Dutch East Indies 36 Dutch learning see rangaku; loans esp. 28, 35-42, 86·--88, 111-··-112, 122, 170···171 (see also DONOR WORD INDEX)
E
G
emuet 90 Early Middle Chinese 160 Economy and Finance Advisory Council 201 Edo 24, 30, 36-37, 42, )"1, 54 Education Law 54 elision 141 ellipsis 22, 130···133, 143, 148-149 English language teaching 57-58; loans esp. 1-3, 11-13,25-28, 42. 50,53-61, 68, 78-;9, 97, 102-1<>5, 112-121, 141-150, 153-15 7> 173-182,185-191,194-195. 2oo---2o6 (see also DONOR
GA; see General American Gaelic loans 112 gaikokugo 8-16, 25, 31, 55, 78, 163, 172-173> 181, 184, 186, 188,195-197,201-202 gairaigo see loans gakusei see Education Law geminates 9, 73, 94, 112---123; see also mora obstruent gendai kanazukai 161, 163, 174 General American 75, 96-97, 102-104, 114, 127, 139, 162,164 gengo tagense tsu 40 Genkai 18, 27 Georgia 68 German loans esp. 28, 40, 50···53, 8;, 98·-·99· 105, 108---109, 122; see also
WORD INDEX)
epenthesis, of the mora obstruent 3, 21, 112---123, 131; of vowels 21, 55, 79, 105-112, 129, 131, 135 Esperanto 173 Etorofu see Iturup euphemism 154 156 Ezo see Hokkaido
11 Feiqe no Afonogatari 30 film titles 11, 13 flaps 72-73> 91 folk etymology 40 foot, prosodic 146 fore-clipping 130-132, 135, 152 foreign advisors 54 57, 79, 97, 102, 12.6, 150 foreign stratuni, definition of ?···14 France 43, 48, 55 Frandscans 29 Fremdwort n Frend1loans esp. 28, 47-50, 89-91, 98-99, 110-111, 121-122; see also DONOR WORD INDEX
fricatives 73, 82-88, 92, 109···-110, 113, 115, 119··-121 Fukuoka 24, 66 Fukuzawa Yukichi 4, 37, 42> 172 full clippings 144, 147 Furansu Jihan 48 fiisetsugaki 36
DONOR WORD INDEX
Germany 43, 51, 55 gerund 141 gijogo see psychomimes giongo see phonomimes giseigo see phonomimes gitaigo see phenomimes glides 74---;6, 92---93, 97···100, 103-104, 161-166, 174, 176,182 go-on 62-63 Golovnin, Vasilli Mikhailovich 45 Great Britain see UK Great Dictionary of the National Language; see
Nihon Kokugo Da~iiten Great East Asia Conference 56 Great Powers 23 Greater East Asia Co-Prosperity Sphere 56 Greek alphabet 159
H Hakka 63 Hakodate 42 half clippings 144, 147 Han 6, 159; Dynasty 6 handakuten 161,169,184 Hang7J1ou 62 Hansen, Gerhard 42 Harris Treaty 42 hatsuontenki 64---66
Subject index 273
Hearn, Lafcadio 54
hebonshiki see Romanization I-Ieike no Monogatari see Feiqe no M~mogatari hentaigana 161 Hirado 2453 himgana 159-163, 166-171, 185,188 Hirohito see Showa Emperor Hiroshima 24,66 Hokkaido 2, 24, 45, 47, 54, 66 Holland 43, 48, 53 homonymy 136 homophony 145, 154 Hong Kong 61, 166 Hoshina Koidli 173 hostile language 56 hybrid stratum Lt··-18; see also compom1ds, hybrid hyphens 177, 18:z,--183 hypocoristic formations 146
I Iberian loans see Portuguese loans, Spanish loans Ichikawa ltsuki 51 ideophonic vocabulary see mimetic stratum imperialism, Japanese 25; linguistic 66 India 30, 44 interpreters, hereditary 36, 45, .f8,54 Italian loans esp. 41 91; see
also DONOR WORD INDEX
Italy 43 iteration marks 163 Ito Hirobumi 43 Iturup 24, 45, 47 Iwakura Mission 43
J Japan Exchange and Teaching Programme 57 Japan Licensing Authority 180 Japan Newspaper Publishers & Editors Association 180 Japanese Empire 5, 61; Foreign Ministry 65; Supreme Court 65 Java 30 jazz argot 147
Jesuits see Jesus, Society of Jesus, Society of 29, 169 JET Programme; see Japan Rxchange and Teaching Programme jiongo; see Sino-Japanese stxatum
Jutsugo I-Iyoki Godo Bukai see Technical Terms and Orthography Joint Subcommittee
K Kaempfer, Engelbert 36 Kagoshima 24, 29 Kaitai Shinsho 36 kan-on 62-63 kana 146, 148, 151, 159-16_3, 169, 171-184, 197; see also
hentaigana, lriragana, katakana Kanagawa Convention 42; Prefecture 42 kango see Sino-Japanese ~traturn
kango yakugo see calques, Sino-Japanese
kanji see Chinese characters Kaiafuto see Sakhalin katakana 7, 28---29, 108, 153, 159-174, 177-181, 185-190
katakanago 7, 195·--198; see also loans Katayarna Toranosuke 201 Keio University 42 Kluner loans 30, 152 Khvo~tov, Nikolai Aleksandrovich 45, 47 Kinki 24 30 kinshipterms 5 kirishitan shiryo see Catholic materials Kitaezo see Sakhalin Kobe 24 42, 55 Koizumi Jun'ichiro 201 Kojien 27 Kojiro Tlmesuke 172 Kokugo Bunkakai see National Language Subcommittee Kokugo Shingikai see National Language Council Kokugoka see National Language Division
Kokurii>;u Kokugo Kenkyiljo see National Institute for Japanese Language
kotJShugo see compounds, hybrid; hybrid stratum Korea 8, 30, 36, 44 66; North 8, 65-67; South 8, 65-66 Korean loans esp. 65-67, 84 123; see also DONOR WORD INDEX
Kulmus, Johann 37 kun'yomi 167; see also semantograrns kunreishiki see Romanizati.on, cabi.r1et-ordered ~tyle Kuomintang 56 lcurikaeshifugo see iteration marks Kurile Islands 45; see also Iturup Kushunkotan 24 45, 47 KY Japanese 191 Kyushu 24, 30-31 I,
labial glides 104, 176 language contact 1-3 Language Normalization Committee 201 language policy, Japanese government 173> 180 laryngealization 83 Late Middle Japanese 30 laterals 91 Latin loans esp. 29··-32, 43; see
aiso DONOR WORD INDE.'i: Lehnwort n Leipzig, Battle of 48 length marks see choonpu lexical stratification 4 Liaodong Peninsula 55 Lippe 36 loan translations 6, 30, 44> s6; see also calques, Sino- Japanese loans, attitudes towards 193--,zo6; auditory 55, 6:z,--
274 Subject index
orthographic 62---63, 67, 76---80, 126; orthography of 159---191; phonology of 71·---136; recognition of 22, 193---195, 206; semantics of 153-155; spelling 67, 79-80, 83, 100, 105, 1J4,
see aiso foreign stratum; Fremdwort; gaikokugo; Lehnwort
122;
Loanword Committee 22, 44, 201-206 London 124 Lyman, Benjamin 54, 150 Lyman's Law 150-151 lyrics, pop n, 16, 20, 159, 186
M Macao 61, 166 MacDonald, Ranald 54 magazines 11,14--21,25,27, 56·-59, 137, 177, 180, 186·---188 Mainichi Shi1Jbun 15, 60-61, 181190 Malacca 29 Manchu 63 Manchukuo 44 manga 163, 166 Matsutoya Yumi 16 medicine, study of in Japan 36---37, 50---51; see
also mngaktt Meiji Era 37, 43, 51, 54, 150; Restoration 21, 42-45, 48, 51,59 merchants, Chinese 31 35; Dutch 21, 23, 33---36, 47; English 53; Portuguese 33 metric system 28, 48,50 mid-clipping 131--·132, 135 militarism 25 mimetic stratum 4-8, 14-18, 75> 146,167 Ming Dynasty 33, 36 missionaries 1123, 29-30, 33, 124, 159, 169; see also Dominicans; Franciscans; Jesus, Society of mixed languages 77 monomoraic articulations 75-76, 92, 184; spellings 176, 179, 182
monophthongization 169 mora 21, 47, 74---76, 84, 8;---88, 94, 110-112, 121, 126-135, 138, 144-150, 161, 163-164, 169, 1jl, 176, 178, 184 202, 204; -clippings 21, 39, /1, 74> 130-132, 135-136, 143, 145, 147; nasal72-75, 89·--90, 94, 106--·107, 127, 145, 161, 174, 178, 181-182; obstruent 3, 21, 73---75, 81, 106, 112-123, 127, 146, 161, 174 181-182; splitting 148 Mori Arinori 54; Ogai 78 Moriyama Einosuke 54 movie titles see film titles multilingualism 1 Muromachi Era 8·-·9
N iN/ see mora nasal Nagasaki 24, 30-3136-38, 42, 45-48, 53-54 66 nakaguro 177, 182-183 Nakasone Hirofumi 68 names, band u, 186; brand 11-12, 23, 56, 68-69, 166, 180, 186; corporate 68, 166, 186, 188, 191; personal 64-66, 123, 180; place 23, 25, 29, 33, 38, 61-68, 81, 85,166 nanbansen 29 Nar*ng 62 Na.ra 5 narrowing, semantic 22, 47, 154,202 Narutaki Juku 37 nasals 72···75, 81, 89---91, 94, 99-101, 106-107, 127, 145, 161, 174> 178, 181-182, 184 National Institute for Japanese Language 4, 14-15, 20, 177···180, 201 National Language Council 173, 175, 177, 180 National Language Division 193 National Language Subcommittee 180 native stratum 4-5, 8-9, 13-18, 75, 128, 138, 146, 151···154>
160, 166--·167, 173, 186--·191, 202 nativization see adaptation Natsurne Soseki 171 neologism 27, 31 79, 168 Netherlands see Holland newspapers 15-20,159-160, 180, 183, 186, 198, 200, 206;
see also Asahi Shi1Jbun; Mamichi Shi1Jbun; Nihon Keizai Shinbun; Yomiuri ShitJbun NHK 65-66, 126, 166, 180, 183, 191, 195-199 l\'ihon Bunten 30 l\'ihon Hoso Kyokai see NHK l\'ihon Keizai Shinbun 193 l\'ihon KDkugo Daijiten 8 Niigata z.t, 42 NJNJAL; see National Institute for Japanese Language Nippo Jisho 30 Nishi 7..enzaburo 37
110bashibo see choonpu Norman Conquest 7; French 7 nouns 1---2, 21, 31, 52, 58, 60---61, 116, 127-128, 137-139, 143, 151, 18o; apophonic 151 novels 16-17, 25, 55, 102 171 nurnerals 1, 5, 159, 190
0 Obuchi Keizo 58 odoriji see iteration marks Ogasa\vara Islands see Bonin Islands Ogata Koan 37 Okinawa 24 57, 63, 76, 101 Okinawan see Ryuk:yuan Okue-zo see Sakhalin Old Japanese 115, 160 on'yomi 167; see also phonograms
onbiki see choonpu onomatopoeia see mimetic stratum
Ontleedkundige 1afelen 37 Orthodox Church 29 orthographic contac.t 3 Orthography Subcommittee 173 Osaka 24 37, 42, 66 Osumi Islands 24> 29
Subject lndex 275 Otorl Keisuke 37 a;•atoi gaikakujin see foreign advisors p
particles 1.-2, 5, 31, 127, 162 past parti.ciples 1.41 Pearl Harbor 56 pejoration 22, 154 Peregrifrafam 29 Perry, Commodore Matthew 42, 54 Phaeton Incident 53 phenomimes 6 phonaesthetic vocabulary see mimetic stratum phonemic analysis, consen-ative 71-75, 82, 86-88, 110, 161, 163; contemporary 72-76, 82-8;1, 86-88, 92-94> 110, 115, 161, 186
phonograrns
73-··76, 84··85, 93. 130, 175··· 179 (see also phonemic
analysis, contemporary) Protestant Churdl 29 psychomimes 6 public language 58 punctuati.on 1.59, 177
Q iQ/ see mora obstruent Qlng 23,36
R
rangaku 6, 35-42. 51 reanalysis 141 Received Pronunciation 93, 96-99,102-105,114
reduction, morphologica.l 21, 134. 137. 141-143
reduplication 6, 146 Renais.~ance
rendaku see sequential voicing replication, anticipatory 111-112
160-161, 170-171
phonomimes 5-6 phonotactics 7···8, 47, 106, 113, 153-154
pidgins .55, 57, .59
Pinto, Fernao Mendes 29 pitdl see accent, Japanese plosives 73, 82-84- 90, 97, 107, 110, 113-116, 119-120, 124-125; re.iilforced 84 plurals 22, 31, 84, 110, 140-143 port lingo 55, 78 portmanteaus 22, 131, 143-144. 149···150 Portuguese loans esp. 27···35, 111-··112, 123, 168···170; see
also DONOR WORD INDEX possessives 22, 141 pre-nasal shortening 99-101, 145
prefixes ;, 58, 139-140 prenasalized consonants 91 prestige 3. 6, 23, 36, 44-45, 54. 61,66,77,102,200
pronouns, personal 1 pronunciation, nonstandard 55; conservative 86, 115, 175 (see also phonemic analysis, conservative); innovative
6
Re7.anov, Nikolai Petrovich 45, 47 Rishiri 24, 54
Roman alphabet
11, 13. 22, 30, 56, 61, 130-131, 159, 169, 185-191, 199 Romani.zation, 173; cabinetordered style 186; hefJilnshiki 124 186 1,
RP see Received Pronunciation ruby 168, 171 Russia 23, 43--45, 55, 68, 167; see also USSR Russian loans esp. 28, 44••"47> 93···94. 109, 122-123; see also DONOR WORD INDEX
Russo-Japanese War 45 Ryukyuan 2, 23,166
s Saint Petersburg 45 Sakhalin 24 37, 44··47, 76 sakoku 23,33.36,40,42 Sano, Tsunetami 37 Sanshiro 171 Sanskrit 6-10 Santa Fe, Paolo de 29 Satta Kotoji 173
Scllambergen, Caspar 36 schwa 79, 90, 102···106 seiyo period 43 semantograms 160, 170-171 sequential voicing 22, 63, 143. 146, 150-15:l
set-inclusi.on model see core-periphery model shakuyogo see loans Shandong 63 Shimabara Rebellion 30 Shirnoda City 24 42; Treaty of 45 Shirnonoseki, Treaty of 55 shinkango see calques, Sino-Japanese Shirando 37 Shizuoka 42 shogunate 21, 30, 36, 45-50, 54· 170
Showa FJnperor ;6 Siberia 45 Siebold, Philipp Franz von 36-37, 40, 50-51 Singapore 56 Sino·· Japanese stratum 4···10, 14-18, 33, 41-44> 52> 56, 61···66, 74 us, 138, 146, 151-154. 170, 185, 188-191, 200, 202, 206; see also
calques, Sino-Japanese Sino-Japanese War, First 55; Second 56 Sinosphere 8 slang, student 52, 55 SMS messaging see texting so-in see fO··it: so··on. see t.iHn soTalon see mora obstruent Song Dynasty, Southern 62 Spanish loans; esp. 33-:35, 91; see also DONOR WORD INDEX SR compounds; see compounds, semantically remodelled stiff voice 83 stigmatization 153 suffixes 123, 138, 141 Sugita Genpaku 36 syllabary 159-161 syllables, closed 75, 106, 121; heavy 74-75, 114, 126-130, 133···135. 145 ··146; light
276 Subject lndex
71·-·7_), 126--·130, 133, 135; open 21; superheavy 74-75, 100, 114, 145 syncope see mid-dipping
T Tagalog 2 Taiwan 61, 65, 166 Taiwanese 166 Take be Tsutomu 201 Tanegashima 24> 29 Tang Dynasty 62 taps 81, 91 Technical Terms and Orthography Joint Subcommittee 173 Teclmical Terms Subcommittee 173 Teki Juku 37, 12 tekiseigo see hostile language television 11, 17, 19 66, 78 textbooks, German 86; high school15; junior high school15; Russian 86 texting 163, 166, 191 tilde 163
to-i11 62 to-on see to-in Toki Zenmaro 173 Tokugawa Yoshimune 36 Tokyo 24, 56, 61, 69, 126, 190 (see also Edo); Bunrika University 173; Hibiya Metropolitan Library 173; hnperial University 51; University of Education 173; University of the Arts 173 topic markers 5 translation bureau 37, 51
trills 81, 91 triphthongs 102---104 Triple Intervention 55 tnmcation 4, :21, 130-1Jl, 142-143, 199; see also acronyms, Roman alphabet; compound reduction; mora-clipping tsuzuriji hatsuon see loans, spelling Turkish 1
u UK % 43-44, 54-56, 68, 79, 96-97, 102, 167 unequal treati.es 42 Uraga 24, 42, 54 USA 23, 42--43,50,54--57, 68, 76, 79, 84, 96·-·97,
102., 124,
126, 140, 167 USSR 23,44
v Vashadze, Grigol 68 verbalization 21, 48, 1:10, 137·-139 verbs 1-2, 52, 58-61, 137-138 vowel harmony see replication, anticipatory vowels, checked 113-121; free 113; lax 114; long 98-104, 114, 145> 147, 163, 169, 171, 174, 178, 182, 184; nasal9o, 101; rhotic 79, 102-105, 11J--n6; short 114; tense 99; see aiso diphthongs; epenthesis of vowels; monophthongization; replication, anticipatory; triphthongs
w wago see native strattml wameihyoki 63-66 wasei no kango 6 waseieigo see compounds, assembled; compounds, made-in-Japan; compounds, semantically remodelled World War I 23, 48, 59; II 21, 47.52,57 -wTitten attestations 25, 32-35, 46, 48-53, 67, 102, 123,139 Wu, Kingdom of 62 Wiirz burg 36
X Xavier, Francis de 29 Xi'an see Chang'an y Yajiro 29 Yale transcription 160 Yamato vocabulary see native stratlml yamatokotoba see native stratlml yeismo 81 yiigo see loans
i'ogo Tekiseika Iinkai see Language Nonnalization Committee Yokohama 24, 42, 55.59 Yorniuri Shinbun 17, 183
z 7..en 62