Handbook of Japanese Phonetics and Phonology

Haruo Kubozono

• Publisher: De Gruyter Mouton
• Year: 2015
• ISBN:
  9781501500596
• Edition: 1
• Topic: Sprachen & Linguistik
• Subtopic: Sprachwissenschaft

This volume is the first comprehensive handbook of Japanese phonetics and phonology describing the basic phonetic and phonological structures of modern Japanese with main focus on standard Tokyo Japanese. Its primary goal is to provide a comprehensive overview and descriptive generalizations of major phonetic and phonological phenomena in modern Japanese by reviewing important studies in the fields over the past century. It also presents a summary of interesting questions that remain unsolved in the literature.

The volume consists of eighteen chapters in addition to an introduction to the whole volume. In addition to providing descriptive generalizations of empirical phonetic/phonological facts, this volume also aims to give an overview of major phonological theories including, but not restricted to, traditional generative phonology, lexical phonology, prosodic morphology, intonational phonology, and the more recent Optimality Theory. It also touches on theories of speech perception and production.

This book serves as a comprehensive guide to Japanese phonetics and phonology for all interested in linguistics and speech sciences.

III Morphophonology

dp n="292" folio="252" ? dp n="293" folio="253" ?

Akio Nasu

6 The phonological lexicon and mimeticphonology

1 Introduction

The vocabulary of Japanese consists of different morpheme classes, each of which has a different etymological origin. As noted in a number of previous works (Martin 1952; McCawley 1968; Vance 1987; Shibatani 1990; Nishio 2002; Ito and Mester 1995a, 2003; among others), at least three classes have been distinguished traditionally: native, Sino-Japanese, and foreign. These three classes can be treated as two large groups as well. One is the native morpheme group, containing items which are indigenous to the language. Items in the native class are called Wago or Yamato.¹⁰⁸ The other group is the loanword class, which consists of Sino-Japanese and foreign vocabulary items. This type of vocabulary developed historically through the process of borrowing from other languages: the Sino-Japanese vocabulary is composed of roots that have been borrowed from Chinese, whereas the foreign vocabulary contains a large number of loans, most of which have been borrowed from European languages.¹⁰⁹

In addition to the morpheme classes mentioned above, there is one more distinctive group of words in the Japanese vocabulary. This is the mimetic vocabulary, which consists of a rich variety of sound-symbolic items. Mimetic words express sounds of the external world in an imitative manner or symbolize states of objects, manners of movement, mental conditions, and so on. In this respect, mimetics can be treated as a special morpheme class which should be distinguished from the other morpheme classes. However, it must be noted that mimetic items are of native origin etymologically; most mimetic words have become established in the Japanese lexicon without any borrowing process. Due to their etymological status, the linguistic treatment of mimetic items has been controversial. On the one hand, some researchers regard mimetic words as a kind of native item. Labrune (2012: 13–14), for instance, remarks that mimetic words “belong to the Yamato class in the strict sense, even if they display a number of properties which may lead one to categorize them in a specific subclass.”¹¹⁰ On the other hand, in several theoretical accounts of the Japanese phonological lexicon, mimetics are treated as an independent lexical stratum from Yamato (Ito and Mester 1995b; Fukazawa 1998; Fukazawa, Kitahara, and Ota 1998, 1999, 2002).

This chapter discusses the phonological properties of the Japanese lexicon with special attention to the peculiar relationship between Yamato and mimetics, in particular with respect to their phonological discrepancies and affinities. In the next section, we will make a general survey of morpheme classes in Japanese and of fundamental ideas about the phonological stratification of the lexicon. In section 3, we will examine a few stratum-specific phonological phenomena with a view to explicitly showing the phonological peculiarities of mimetics. In section 4, we will review the theoretical models presented in the previous literature on lexical stratification. Section 5 discusses the status of mimetics in the phonological lexicon of Japanese.

2 Stratification of the lexicon

2.1 Etymological classes and phonological properties

It is well known that morphemes belonging to distinct etymological classes exhibit different phonological properties from each other. Yamato items are distinctive in that they are subject to a well-known compound voicing process called rendaku (Martin 1952; Hamada 1960; Nakagawa 1966; Sakurai 1972; Kindaichi 1976; Vance 1987; Satō 1989; among others). In this process, the initial voiceless obstruent of the second element of a compound is voiced (e.g., tori ‘bird’ > oya+dori ‘parent bird’). It occurs frequently in Yamato stratum, but loanwords do not undergo the process in general; see Vance (this volume) for a detailed discussion. In addition to rendaku, Yamato has a conspicuous restriction with respect to voicing; voiced obstruents (i.e., /b, d, z, g/) generally do not occur underlyingly in morpheme-initial position. Voiced obstruents are favored in morpheme-medial position in Yamato items such as haba ‘width’, kuda ‘tube’, kaze ‘wind’, toge ‘thorn’ (Hashimoto 1938; Komatsu 1981; NLRI 1984; Ito and Mester 1986; Labrune 2012; and see also Takayama, this volume, for a historical discussion).

Sino-Japanese items are characterized in terms of their unique shape. Sino-Japanese roots are in principle monosyllabic and variations of the syllable structure are limited to only the following four types: CV (ka ‘course, department’), CVV (doo ‘copper’, suu ‘number’, zei ‘tax’, kai ‘meeting’, rui ‘sort, class’), CVN (kin ‘gold’), or CVCV (betu ‘distinction, other’, koku ‘nation’).¹¹¹ Though the last pattern has disyllabic structure, its underlying form is interpreted as monosyllabic /CVC/ and an epenthetic vowel (/u/ or /i/) is attached to the coda consonant to prevent a closed syllable from emerging (see Ito and Mester 1996 as well as Ito and Mester, this volume, for further discussion). In addition to monosyllabicity, palatalization of the onset consonant is characteristic of Sino-Japanese (Nakata 1982: 308–311). That is, Sino-Japanese contains a number of syllables in which the onset is palatalized, such as tya ‘tea’, kyuu ‘emergency’, and myoo ‘strange, mystery’. McCawley (1968: 62–66) is an early theoretical attempt to account for the phonological diversity among morpheme classes with respect to the distribution of palatalized (“sharp” in his terminology) consonants.

The foreign stratum has many characteristic properties that distinguish the items involved from those in other strata. First, the emergence of voiced geminates, as in beddo ‘bed’, is a conspicuous feature of the foreign stratum (Martin 1952; Ito and Mester 1995a,b; Katayama 1998; Irwin 2011; Labrune 2012; among others). Second, the voiceless bilabial stop [p] can freely appear as a licit surface segment in foreign items such as paipu ‘pipe’, puuru ‘pool’, supai ‘spy’, etc. (McCawley 1968: 77–85; Shibatani 1990: 166–167; Ito and Mester 1995a,b; Labrune 2012: 70–77; among others). Third, foreign items frequently contain novel CV sequences which do not appear in Yamato and Sino-Japanese words (Hattori 1979; NLRI 1990; Ito and Mester 1995a,b; Katayama 1998; Irwin 2011; among others). For example, syllables containing the voiceless bilabial fricative [ɸ] can appear with no restrictions in foreign words such as [ɸ]aito ‘fight’, [ɸ]iibaa ‘fever’, nai[ɸ]u ‘knife’, ka[ɸ]e ‘café’, and [ɸ]ooku ‘fork’, whereas it can appear only before /u/ in Yamato and Sino-Japanese. NLRI (1990: 62–74) and Irwin (2011: 75) present the list of CV moras found only in foreign items; see also Pintér (this volume) and Kubozono (this volume).

2.2 Phonological stratification

As long as we restrict our attention to the facts mentioned above, each etymological class seems to have its own phonological properties that characterize that class exclusively. However, it is not always the case that a phonology-based characterization of lexical strata directly corresponds to the etymological classification of lexical items in a one-to-one fashion. Some phonological properties are shared among two or more morpheme classes. The following data, for example, show that items in etymologically different morpheme classes pattern together. In (1)–(2) and the rest of the chapter, dots (.) denote syllable boundaries.

The voiceless bilabial stop [p] cannot appear as a syllable onset following a vowel either in Yamato or in Sino-Japanese. In these two classes [p] is converted to [h], as exemplified in (1). With respect to the illegitimacy of [p], Yamato and Sino-Japanese pattern together. On the other hand, the data in (2) show that another grouping can be established, from which Yamato is excluded. An obstruent in postnasal position can be voiceless both in Sino-Japanese and in foreign items, while it must be voiced in Yamato forms such as tonbo ‘dragonfly’, sinda ‘died’, kangae ‘thought, idea’, etc. On the basis of the phonological patterns in (1) and (2), the following two groupings can be established for these morpheme classes.

The phonology-based classification in (3) implies that the etymological partitioning of morpheme classes cannot account for all the properties of the synchronic configuration of the lexicon. The synchronic lexicon is, rather, organized gradiently, with some phonological properties overlapping between two (or more) morpheme classes.

With reference to the phonological classification of lexical items, it is notable that mimetic items exhibit quite distinctive behavior with respect to the phonological regularities discussed above. Although mimetic items are of native origin etymologically, they are exempt from the prohibition against singleton [p]; it appears as a licit segment, as exemplified in (4a). Moreover, as shown in (4b), mimetic items are subject to the process of postnasal voicing, just like Yamato items. (“-” denotes a morpheme boundary.)

These data indicate the dual character of mimetics. While mimetics behave as if they belong to the native stratum with respect to the postnasal voicing (4b), they are the opposite of Yamato with respect to the legitim...

Table of contents

• Handbooks of Japanese Language and Linguistics
• Title Page
• Copyright Page
• Preface
• Introduction to the Handbooks of Japanese Language and Linguistics
• Table of Contents
• Contributors
• I Introduction to Japanese phonetics and phonology
• II Segmental phonetics and phonology
• III Morphophonology
• IV Prosody
• V Broader perspectives
• Subject index