第10号 (2016年1月)

国立国語研究所論集 / NINJAL Research Papers
第10号 2016年1月 / No. 10 January 2016

ISSN: 2186-134X print/2186-1358 online

目次 Contents | PDF (529KB)


Yilan Creole Case Marking

CHIEN Yuehchen
簡 月真

抄録 Summary | PDF (639KB)

National Dong Hwa University, Taiwan / Project Collaborator, NINJAL [-2013.09]


Yilan Creole is a Japanese-based creole language spoken in Taiwan. It has been used as a first language by indigenous Atayal and Seediq people living in Yilan County. This paper is a descriptive study that attempts to clarify the linguistic nature of Yilan Creole through an examination of its case-marking system. Both word order and postpositions are used for case marking in Yilan Creole. Basically, the subject and direct object are indicated by word order, while the indirect object and other arguments are marked by five postpositional case particles: ni, de, to, no, and kara. The case particles are derived from the lexifier language, Japanese. The paper concludes that, compared to the Japanese case particles, the case particle ni has undergone an extension of functions, and some other particles, on the contrary, saw a reduction of functions in Yilan Creole.

Key words: Taiwan, Yilan Creole, Japanese, Atayal, case marking

台湾国立東華大学/国立国語研究所 共同研究員[-2013.09]


宜蘭クレオールは台湾で話されている日本語を語彙供給言語とするクレオール語である。台湾東部の宜蘭県においてアタヤル人及びセデック人の第一言語として使われているが,若い世代では華語へシフトしつつあり,消滅の危機に瀕している。本稿は,この言語の格表示に焦点をあて,その特徴を記述するものである。宜蘭クレオールでは,語順及び後置詞を格表示として用いている。具体的には,主語と直接目的語は語順,間接目的語とその他の項は5つの格助詞「ni, de, to, no, kara」によってマークされている。これらの格標識は上層言語である日本語由来のものであるが,そこには異なった用法が存在し,単純化への変化が認められる。また,niの意味用法の拡張なども見られ,独自な格表示のシステムが作り上げられている。



Reanalysis of the Meaning of ko-: The New Minimal Semantic Contrast in Demonstrative Systems


抄録 Summary | PDF (1,090KB)

秋田大学/国立国語研究所 共同研究員




Akita University / Project Collaborator, NINJAL


This paper provides an analysis of the Japanese demonstrative pronoun ko- in its deictic use, based on spontaneous interaction data. The concept of new minimal semantic contrast introduced by Levinson (2000) and applied by Enfield (2003) is used as the framework of analysis. In contrast to previous accounts of ko- and a-, in which ko- encodes [+proximal] and a- encodes [-proximal], distributional findings show that the indicating ranges of ko- and a- are asymmetrical. While a- indicates only referents that are in a certain location, ko- has flexibility with regard to the spatial location of its referents and is also conceivable in contexts where a- is used. The present discussion reasons that ko- does not encode any spatial information, and shows that the proximity of ko- is not determined by semantic meaning, but by conversational implicature. This reasoning corresponds with the nature of the Lao demonstrative determiner system analyzed by Enfield (2003).

Key words: deixis, demonstrative, conversational implicature, unmarked demonstrative form, engagement area


Long-term Changes of Functional Elements in the Okazaki Survey of Honorifics: Focusing on Interpersonal Consideration in a "Luggage Leaving" Scene

丁 美貞
JEONG Mijeong

抄録 Summary | PDF (976KB)

国立国語研究所 時空間変異研究系 非常勤研究員





Adjunct Researcher, Department of Language Change and Variation, NINJAL


This paper describes the results of analyzing the functional elements of data from the "luggage leaving" scene in the Okazaki survey of honorifics, one of the long-term studies by the National Institute for Japanese Language and Linguistics (NINJAL).

The analysis is informed by the research of Kumagai and Shinozaki (2006), who classify the reaction statements in the "luggage leaving" scene into communication functions and functional elements. In the scene, there are six communication functions, i.e., "starting a conversation," "briefing," "effective reinforcement," "interpersonal consideration," "prompting action," and "others." In addition, there are 10 functional elements of subdivision, i.e., "calling for attention," "business," "circumstances," "inconvenience," "assurance," "request for leaving luggage," "reminding of the request," "confirmation of the intention," "expressing gratitude," and "others." This paper presents usage examples of the communication functions and the functional elements, their frequency of use, and long-term changes. Furthermore, it examines the long-term changes of interpersonal consideration on which it is focused.

Key words: functional elements, communication functions, interpersonal consideration, expressing gratitude, generation gap


Colloquial Japanese in Materials from the Early Modern Period: Shorthand Writings and 78 rpm Disk Recordings of Rakugo in the Kanto and Kansai Regions


抄録 Summary | PDF (1,031KB)

横浜国立大学/国立国語研究所 共同研究員




Yokohama National University / Project Collaborator, NINJAL


This article compares shorthand writings of rakugo (comical stories) published during the Meiji Period and their spoken versions recorded on 78 rpm disks during the same period. It examines characteristics of language use in the two media, particularly focusing on identical stories by the same storytellers.

Although similarities are to be found in the structure of the stories and lexical choice, they bear little resemblance in terms of the characteristics of spoken language. The findings suggest that it is necessary to investigate the characteristics of shorthand writings as a linguistic resource, before we make use of them for research as representative of spoken Japanese in the Early Modern Period.

Key words: rakugo (comical story), shorthand writing, 78 rpm disk record, characteristics of Japanese spoken language


Survey of Conversational Behavior: Towards the Design of a Balanced Corpus of Conversational Japanese

小磯花絵a,土屋智行b,渡部涼子c,横森大輔b,相澤正夫d,伝 康晴e
KOISO Hanaea, TSUCHIYA Tomoyukib, WATANABE Ryokoc, YOKOMORI Daisukeb, AIZAWA Masaod and DEN Yasuharue

抄録 Summary | PDF (1,490KB)

a国立国語研究所 理論・構造研究系
c国立国語研究所 コーパス開発センター 非常勤研究員
d国立国語研究所 時空間変異研究系
e千葉大学/国立国語研究所 言語資源研究系 客員教授




aDepartment of Linguistic Theory and Structure, NINJAL
bKyushu University
cAdjunct Researcher, Center for Corpus Development, NINJAL
dDepartment of Language Change and Variation, NINJAL
eChiba University / Invited Professor, Department of Corpus Studies, NINJAL


In this paper, we report on the activity of a preparatory project to build a large-scale corpus of conversational Japanese (NINJAL collaborative research project, 2014/7/1-2015/8/31). The overall aims of this project are: i) to establish a corpus design for collecting various kinds of everyday conversations in a balanced manner, ii) to develop a methodology of recording naturally occurring conversations, and iii) to create a transcription system suitable for effectively transcribing natural conversations. This report focuses on the first issue of establishing a corpus design. We first describe our survey of everyday conversational behavior, conducted with about 250 Japanese adults last year, in order to reveal how diverse our everyday conversational behavior is, and to build an empirical foundation for corpus design. The questionnaire included when, where, how long, with whom, and in what kind of activity informants were engaged in conversations. We found that ordinary conversations show the following tendencies: i) they mainly consist of chats, business talks, and consultations; ii) in general, the number of participants is small and the duration is short; iii) many conversations are conducted in private places such as homes, as well as in public places such as offices and schools; and iv) some questionnaire items are related to each other. Based on these results, we discuss how to design a balanced corpus of conversational Japanese.

Key words: survey of conversational behavior, corpus design, corpus of conversational Japanese


A Qualitative Analysis of the Diversification of Teaching Values of a Korean Vocal Trainer Working in Japan

李 奎台
LEE Kyutae

抄録 Summary | PDF (878KB)

東京外国語大学大学院 博士後期課程/国立国語研究所 日本語教育研究・情報センター 非常勤研究員[-2015.03]




Doctoral Student, Tokyo University of Foreign Studies / Adjunct Researcher, Center for JSL Research and Information, NINJAL [-2015.03]


This paper considers how the teaching values of a Korean vocal trainer working in Japan have changed over time. The longitudinal observation and interviews were conducted over a period of 1 year and 3 months after she began working in Japan. The data were qualitatively analyzed based on SCAT (Steps for Coding and Theorization). Upon starting work, she was confronted with a high dropout rate among her students. Despite her attempts to reduce this, the dropout rate rose. In this critical situation, she consulted with other trainers and with her students. Through such consultation, she came to realize others' values, which led her to reflect upon her own teaching practices. Subsequently, her teaching values appeared to diversify. It is through this process of diversification that she became aware of the ways in which her former teaching had been inappropriate. Thus, the main factor leading to diversification of her teaching values appears to have been the self-reflection following interactions with others, which had been prompted by the critical situation itself. This paper reveals the process of the diversification of values caused by a critical situation and its consequent problem-solving.

Key words: career formation, foreign workers, teaching values, SCAT, self-reflection


What Does Compound Accentuation Tell Us about the History of Japanese?


抄録 Summary | PDF (868KB)

日本女子大学/国立国語研究所 時空間変異研究系 客員教授[-2013.03]




Japan Women's University / Invited Professor, Department of Language Change and Variation, NINJAL [-2013.03]


This paper argues that the compound accentuation rule in which the accent of the "first" member of compounds is preserved is the most archaic type in the Japanese and Ryukyuan dialects. By contrast, other types, such as the one in which the accent of the "second" members are preserved, or the one by which the default accent is placed around the boundaries of the two members, are newly developed compound accentuation rules. The paper then argues that even in Tokyo Japanese, in which productive compound accentuation is decided exclusively by the "second" members, we find some vestiges of the older compound accentuation rule in which the "first" member of the compounds is still relevant.

Key words: compound accentuation, connecting accent, Japanese dialects, Ryukyuan languages, Yamada's Law


Phonological and Phonetic Description of Voiced Geminates in Amakusa Japanese


抄録 Summary | PDF (1,206KB)

北星学園大学/国立国語研究所 共同研究員




Hokusei Gakuen University / Project Collaborator, NINJAL


This article reports on the phonological distribution and acoustic phonetic realizations of voiced geminates in Amakusa Japanese. Results from fieldwork elicitation demonstrate that most dialects in Amakusa Japanese have voiced geminates, not only in foreign words but also in Yamato and Sino-Japanese words. Moreover, the findings indicate that the fully-voiced pattern in geminates is found at many closure intervals.

Key words: voiced geminates, Amakusa Japanese, closure voicing


Classification of Japanese Compounds Based on the Frequency of Rendaku: A Study Using the Rendaku Database

太田 聡a,太田真理b
OHTA Satoshia and OHTA Shinrib

抄録 Summary | PDF (809KB)

a山口大学/国立国語研究所 共同研究員


連濁はもっとも広く知られた日本語の音韻現象の1つである。先行研究では,日本語の複合語は連濁の生起率の違いに基づいて,いくつかのグループに分類されることが提案されている。しかしながら先行研究では,連濁生起率の分類基準が恣意的であった点,またグループの数をあらかじめ仮定していた点に問題があった。そこで本研究では,混合正規分布モデルに基づくクラスター分析と連濁データベース(Irwin and Miyashita 2015)を用いて,日本語複合語を分類する際の最適な分類基準とクラスター数を検討した。複合名詞と複合動詞のどちらも,2つのクラスターを仮定したモデルが最適であり,クラスター同士の分類基準は,複合名詞では連濁生起率が90%,複合動詞では40%であった。これらの結果は先行研究のクラスター数や分類基準とは異なるものであった。我々の結果は,モデルに基づくクラスター分析が言語データに対する最適な分類を行う上で非常に有効であることを示すものである。


aYamaguchi University / Project Collaborator, NINJAL
bThe University of Tokyo


Rendaku is one of the most well-known phonological phenomena in Japanese, which voices the initial obstruent of the second element of a compound. Previous studies have proposed that Japanese compound words can be classified on the basis of the frequency of rendaku (rendaku rate). However, since these studies used arbitrary criteria to determine clusters, such as 33% and 66%, as well as arbitrary numbers of clusters, it is crucial to examine the plausibility of such criteria. In this study, we examined the optimal boundary criteria as well as the optimal number of clusters using a clustering analysis based on Gaussian mixture modeling and the Rendaku Database (Irwin and Miyashita 2015). The cluster analyses clarified that the two-cluster model was optimal for classifying both compound nouns and compound verbs. The boundary values of the rendaku rate for these clusters were approximately 90% and 40% for the compound nouns and compound verbs, respectively. These results were inconsistent with the findings of previous studies. Our findings demonstrate that model-based clustering analysis is an effective method of determining optimal classification of linguistic data.

Key words: rendaku, compound word, frequency, cluster analysis, Gaussian mixture model


The Indirect Interrogative and its Relational Constructions in Late Modern Japanese - A Network between Constructions with the Embedded -ka Clause -


抄録 Summary | PDF (1,088KB)

名古屋大学大学院/国立国語研究所 共同研究員


本研究は,主に高宮(2003,2004,2005)の一連の研究によって明らかにされた間接疑問構文の歴史的な発達について,その痕跡が明治期の日本語にどの程度見られるかを,小説(文学)テクストのコーパスから抽出された用例をもとに,現代語とも対照しながら記述した。間接疑問構文の主節述語は,近代に入っても未だ未決タイプ(「知らない」「分からない」等)が多いが,江戸後期には未発達であった既決タイプ(「分かる」「知っている」等)も1 割を超える割合で現れ,対処タイプ(「考える」「確かめる」等)においても形態的な制約がなくなり,主節述語のヴァリエーションが増えていることが確認された。また,間接疑問節のタイプでは,疑問詞疑問のタイプが非常に優勢であることも明らかになった。



Nagoya University / Project Collaborator, NINJAL


The studies of Takamiya (2003, 2004, 2005) have revealed that the Japanese indirect interrogative construction evolved gradually from the Muromachi (14th-16th C. AD) to the Edo period (17th-19th C. AD). We demonstrate how much of its historical trace can be found in Late Modern Japanese through the analysis of literature corpus data from the Meiji era (1868-1912) and describe its properties of meaning and construction, which are further contrasted with present-day Japanese. We confirm that the literature texts in the Meiji era still contain a high incidence of the main predicate "absence of information type" (not know, not understand, etc.), while the incidence of the "existence of information type" (know, be obvious, confirm, etc.), which did not appear in Early Modern Japanese, exceeds 10%. Furthermore, there is no morphological restriction in the "search for information type" (check, ask, etc.) in Late Modern Japanese, by contrast with Early Modern Japanese.

We presume the embedded construction with the psychological main predicate to be a typical indirect interrogative and deliberate the relationship (network) between other constructions with the embedded -ka clause, such as the dependence, indirect exclamatory, anaphora, concealed question, content clause, and two sentences in succession constructions, showing common and differing properties of their meaning and construction.

Key words: indirect interrogative construction, indirect exclamatory construction, two sentences in succession construction, concealed question, the absence of information type


Regional Differences in the Pain Expression uzuku

TAKEDA Kōkoa and YARIMIZU Kanetakab

抄録 Summary | PDF (1,986KB)

a国立国語研究所 時空間変異研究系[-2015.03]
b国立国語研究所 時空間変異研究系 非常勤研究員








aDepartment of Language Change and Variation, NINJAL [-2015.03]
bAdjunct Researcher, Department of Language Change and Variation, NINJAL


In this paper, we clarify the differences in region, generation, and meaning of the verb uzuku, which is used to express a type of pain. We consider the characteristics of its usage through analysis of data from "The Nationwide Survey for Chronic Pain and its Expressions," which was administered to approximately 180,000 people.

The verb uzuku is used to diagnose the clinical condition of the patient in a medical context. It is regarded as part of the standard Japanese language and is also used in the survey. However, an analysis of the results of this survey found that uzuku is mainly used in western Japan and by people in their 50s or older.

In terms of the differences in meaning, the rate of use declines nationwide in descending order when referring to "toothache," "cut," "headache," "arthralgia," "backache," and "stomachache."

Uzuku is used mainly in western Japan to express "toothache," "cut," and "headache." However, it is hardly used for headaches in eastern Japan, while its rate of use in western Japan also declines in under 50s. In Ehime Prefecture, its rate of use is high in expressing "arthralgia," "backache," and "stomachache." Therefore, uzuku has a broader meaning in the Ehime dialect than in other dialects.

With regard to the quality of pain, the type of pain is differentiated to an increasing degree from that of "toothache" in the case of "headache," "arthralgia," "backache," and "stomachache." In the order of continuity of pain from "toothache," the rate of use of uzuku decreases and the area of use becomes narrower.

Uzuku is part of an "unnoticed dialect," which expresses physical sensation and is standardized with reduction of meaning. Physical sensation is personal, and words of physical sensation tend to be used in the private domain. Therefore, it can easily be overlooked that they are dialect forms.

There are many answers to the survey that can be given other than uzuku. These are used differently according to usage or nuance. It is suggested that the meaning of uzuku has become narrower because itai/itamu has taken on the core meaning of pain, especially in the Kanto district.

Key words: The Nationwide Survey for Chronic Pain and its Expressions, continuity of pain, unnoticed dialect, difference in generation, standardization of the Japanese language


Practical Uses for Research Materials Owned by NINJAL


抄録 Summary | PDF (1,949KB)

国立国語研究所 研究情報資料センター 非常勤研究員[-2015.08]




Adjunct Researcher, Center for Research Resources, NINJAL [-2015.08]


Since its inception, The National Institute for Japanese Language and Linguistics (NINJAL) has been accumulating research documents and media for linguistics (e.g., phonetic data recordings on cassette and digital audio tapes). This paper aims to explain the contents of these materials owned by NINJAL. A further important point of consideration is creating a database for these abundant materials. I note the advantages of using a Relational Database Management System (RDBMS) for the description of research materials and the contents of previous research projects, which is an improvement on XML as formerly used in our laboratory.

Key words: archives, EAD, RDBMS, audio-visual data, digital data


Accent Data from the Kikai-jima Dialects: Part 4


抄録 Summary | PDF (734KB)

東京大学名誉教授/国立国語研究所 理論・構造研究系 客員教授[-2015.03]




Emeritus Professor, The University of Tokyo / Invited Professor, Department of Linguistic Theory and Structure, NINJAL [-2015.03]


Accent data from the Nakasato and Sakamine dialects in Ryukyuan Kikai-jima are presented with particular reference to productive compound nouns whose first members are place names. Analysis reveals that a number of words have a distinct γ pattern and that both dialects have a three-pattern accent system.

Key words: Kikai-jima dialects, Nakasato dialect, Sakamine dialect, third pattern (γ pattern), three-pattern accent


Differences in Interactive Usage of Japanese Recipient Response Tokens un and hai


抄録 Summary | PDF (830KB)

早稲田大学/国立国語研究所 共同研究員[-2014.03]


本研究は「うん」「はい」に代表される「聞き手の短い反応」を対象とし,その使い分けを明らかにするものである。これまでの日本語学・日本語教育における研究では,「うん」と「はい」の違いは基本的には「丁寧さ」のみが異なるものとして提示されてきた。しかしながら,筆者が「日本語話し言葉コーパス(CSJ)」のインタビュー場面のデータを対象とし,分析を行ったところ,「丁寧さ」とは異なる基準で両者が使い分けられている可能性が示唆された。本稿では,会話分析(Conversation Analysis)の立場から,インタビュイー(説明者)が自己開始修復を用いて「聞き手に向けて特別に説明を差し挟み,修復の操作が完了する」とわかる位置において,聞き手はそれまでの反応とは差異化した形で反応をすることを例証する。


Waseda University / Project Collaborator, NINJAL [-2014.03]


This study highlights usage differences between un and hai, two recipient response tokens available in the Japanese language. Prior research in the fields of Japanese language and Japanese language education suggests that the difference in usage between the two response tokens is merely one of politeness. However, analysis by the author of the Corpus of Spontaneous Japanese (CSJ), a collection of Japanese conversations, indicates that recipients may also distinguish the use of un and hai based on a standard other than politeness. From the viewpoint of conversation analysis, this paper illustrates that when speakers use self-initiated repair during an explanation to supplement special information for recipients, the latter will switch to a different response token at the end of the repair sequence, thus differentiating between the tokens on the basis of context.

Key words: recipient response tokens, un/hai, interaction, self-initiated repair


e-learning を利用した韓国語母語話者によるライティングの誤用への間接的フィードバックの研究
A Study of Indirect Feedback on E-learning Writing Errors by Korean Learners of Japanese as a Foreign Language

尹 鎬淑a,迫田久美子b,川崎千枝見c
YOUN Ho-Sooka, SAKODA Kumikob and KAWASAKI Chiemic

抄録 Summary | PDF (848KB)

aサイバー韓国外国語大学校/国立国語研究所 外来研究員[-2015.08]
b国立国語研究所 日本語教育研究・情報センター


(i) 複数回の間接的フィードバックは,韓国人日本語学習者の誤用訂正を促進するか。言い換えれば,さらに気づきを促す効果があるか。
(ii) 促進するとすれば,日本語能力の上位群と下位群で効果に違いが見られるか。
(iii) 複数回の間接的フィードバックにより誤用訂正を促進する特定の文法項目(受身,助詞,モダリティ,自他動詞)があるか。

調査の結果,(i)' 1回目より2回目のフィードバックによる訂正率が高く,複数回の間接的フィードバックは,自己訂正を促進することがわかった。また,(ii)' 下位群は上位群に比べて誤用訂正の効果が高いことが明らかになった。(iii)' 文法項目については,学習者は日本語母語話者に比べて,受身やモダリティを過剰使用する傾向が見られたが,フィードバックによって修正ができた。助詞については,1回目のフィードバックでの自己訂正の割合が高いが,2回目のフィードバックの後でも依然として誤用が残るケースがあるという結果となった。このことから,助詞は自己訂正しやすい項目ではあるが,複数回のフィードバックでも誤用が残り,間接的フィードバックによる自己訂正が極めて困難な項目であることがわかった。



aCyber Hankuk University of Foreign Studies / Visiting Researcher, NINJAL [-2015.08]
bCenter for JSL Research and Information, NINJAL
cHiroshima Kokusai Gakuin University


The aim of this article is to examine the three questions below based on the results of written data by Korean learners of Japanese as a foreign language.
i) Do repeated indirect feedbacks (FB1 & FB2) promote self-monitoring more than one-time feedback (FB1)?
ii) Is there a difference between advanced-level learners and elementary-level learners?
iii) Do these forms of feedback promote learners' self-monitoring to grammatical items of Japanese, such as the passive, particles, modality, and transitive/intransitive verbs?

According to the results, i) both forms in the study promote self-monitoring, with the second feedback (FB2) more effective than the first (FB1); ii) compared to the advanced group, the elementary group corrects errors more by repeated feedback; iii) learners overuse grammatical morphemes of the passive and modality by comparison with native speakers of Japanese. Both forms of feedback are very effective in correcting particle errors. However, learners have difficulty in correcting all errors by repeated indirect feedback.

Key words: e-learning, writing errors, indirect feedbacks, self-monitoring, noticing


Two-Chinese-character Words in Missionaries' Documents in China: A Comparison with the Written Documents of Dutch Studies in Japan

朱 京偉
ZHU Jingwei

抄録 Summary | PDF (1,145KB)

北京外国語大学/国立国語研究所 理論・構造研究系 客員教授[-2010.10]







Beijing Foreign Studies University / Invited Professor, Department of Linguistic Theory and Structure, NINJAL [-2010.10]


In a previous article (Zhu 2015), I considered two-Chinese-character words in written materials of Dutch Studies in the Edo era. The current paper focuses on Chinese missionary materials, comparing two-Chinese-character words found therein with those of the Dutch Studies' written documents. My primary conclusions are as follows.

First, with regard to word structure patterns, differences frequently occur in adnominal two-Chinese-character words. In missionary materials, the percentage of N+N pattern is very high, while in the materials of Dutch Studies V+N and A+N are high.

Second, there are 186 common words in both materials, of which 85.5% (159 words) are from classical Chinese and 14.5% (27 words) have a new meaning or are not from classical Chinese. The words from classical Chinese have not been adopted from the materials of Dutch Studies into missionary materials, or vice versa; rather, they have been separately adopted from classical Chinese. On the other hand, the words that are not from classical Chinese fall into one of the three following categories: i) words separately created in China and Japan that coincidentally match in form, ii) words for which an earlier source could not be found due to the limited scope of the study, or iii) words that spread in some way from the documents of Dutch Studies to missionary materials.

Lastly, with regard to the relationship between two-Chinese-character words and three-Chinese-character words, in this study it became clear that the greater the number of neologisms in the posterior morpheme of two-Chinese-character words, the higher the probability that they would be used as the posterior single-morpheme of three-Chinese-character words. On the other hand, two-Chinese-character words and four-Chinese-character words are only weakly connected: in missionary materials, only 3.7% of two-Chinese-character words are used as the anterior morpheme of four-Chinese-character words, and only 5.0% as the posterior morpheme.

Key words: two-Chinese-character word, missionaries' documents in China, written documents of Dutch Studies, word formation pattern, morpheme