Diachrony of Verb Morphology

Japanese and the Transeurasian Languages

Martine Robbeets

• Publisher: De Gruyter Mouton
• Year: 2015
• eBook ISBN: 9783110400113
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Historical & Comparative Linguistics
• Index: Languages & Linguistics

This book deals with shared verb morphology in Japanese and other languages that have been identified as Transeurasian (traditionally: "Altaic") in previous research. It analyzes shared etymologies and reconstructed grammaticalizations with the goal to provide evidence for the genealogical relatedness of these languages.

1 Introduction

1.1 Goals of this study

The question of whether the languages here referred to as Transeurasian, namely the Japanic languages, the Koreanic languages, the Tungusic languages, the Mongolic languages and the Turkic languages, constitute a genealogical grouping remains one of the most disputed issues in comparative historical linguistics. The controversial classification has been on the table for nearly two centuries, but in spite of recent claims from both supporters and critics that the controversy has been resolved (Starostin et al. 2003: 7, Vovin 2005a: 71), the debate is too complex to be easily settled. It is a sign of the healthy state of the Transeurasian debate that it continues to generate new evidence and counter-evidence. The present volume has no intention to close the debate on Transeurasian classification, but it will propose new evidence from verb morphology, in addition to putting old evidence in a new light. Using the traditional comparative method as a basic tool, the present volume examines whether the correspondences in verb morphology between the Transeurasian languages suggest a common ancestorship.

What are the objectives of this study? First, it aims to contribute to the debate about the genealogical relationship of Japanese and to the discussions on the affinity of the so-called “Altaic” languages. The present work starts from the proposition that both issues are so closely interrelated that it is virtually impossible to discuss them independently. This assumption challenges a commonly used starting point in the Altaic literature, i.e. the following observation by Georg (2007: 268): “The core of the “Altaic” debate is the question whether the three (“western”, “continental”) families Turkic, Mongolian and Tungusic, are genetically related to each other in the first place, and only secondarily, whether Korean (“peninsular Altaic”) and/or Japanese (“insular Altaic”) can or should be added to this grouping.” I argue, however, that it is nearly impossible to demonstrate that Turkic, Mongolic and Tungusic are genealogically related without considering Korean and Japanese. This viewpoint is supported by another observation made by the same author (Georg 2003: 433): “…most critics of Altaic maintain that these languages do share a great deal … of… historical identical lexical and morphological elements, but that these -at least their majority- is better accounted for by assuming large-scale mutual borrowing than common genetic descent.” It appears that the question whether the Transeurasian languages are genealogically related essentially boils down to a clash between diffusionist and retentionist explanations. Thus, the primary controversy in the Transeurasian debate is not fuelled by a shortage of similarities, but by the difficulty of accounting for them. The key question is whether all shared forms are generated by code-copying or whether some are the residues of inheritance. In the past, each of the three western families and the two eastern families maintained high-contact relationships amongst themselves. If a shared morphology is found between low-contact languages, including, for instance, Japanese and Turkic, code-copying can be ruled out with high probabilty. Starting from a hypothesis that includes low-contact languages therefore offers the best chance of resolving the longstanding copy-cognate debate for the Transeurasian languages.

In addition to contributing to the distinction between borrowing and inheritance, the present volume will address other objections against the genealogical relatedness of the Transeurasian languages. Ramer et al. (1998:79–88) identify the main issues as the lack of a common basic vocabulary, the establishment of phonological correspondences, the lack of attention to common morphology including the interpretation of the shared personal pronouns and the lack of consistency with the evidence in support of Indo-European. The recent confrontation of opinions in publications such as Starostin et al. (2003), Robbeets (2004, 2005) and reviews of these studies has advanced our understanding of matters relating to phonology and basic vocabulary. The obstacles largely unaffected by these discussions are first, the need to identify a common morphology, second, the problem of distinguishing between code-copying and inheritance without over-reliance on vocabulary and third, the challenge of explaining the un-Indo-European characteristics of the evidence. As these issues are interconnected, treating one issue will elucidate the others. Thus, the present demonstration that the Transeurasian languages share a common morphology will shed light on the distinction between code-copying and inheritance, while insight into the different nature of the shared properties will prevent us from setting Indo-European as a standard against which Transeurasian is to be assessed.

A third objective of the present study is to add to the reconstruction of the common ancestor of Japanese and Korean. Whereas the inclusion of Japanese and Korean will shed light upon the nature of the relationship between the other Transeurasian languages, the reverse is also true; the comparison with Turkic, Mongolic and Tungusic is expected to have implications for the reconstruction of proto-Japanese-Korean. Although the pitch of their arguments mainly rely on lexical evidence, Martin (1968, 1990, 1991a, 1995, 2002) and his student Whitman (1985) have proposed a common morphology in support of a Japanese-Korean unity. Janhunen (2010: 132), however, warns that “Since Japanese and Korean are spoken in adjacent areas, a large proportion of their shared features must, in any case, be due to contacts, and it is important to eliminate the illusion of similarity created by these contacts before progressing to the level of deeper genetic studies.” In line with this caveat, Vovin (2010) indeed attempts to refute the Martin-Whitman etymologies, proposing that the forms were actually copied from Korean into Japanese in early historical times. Unger (2009) attempts to rehabilitate the Martin-Whitman corpus by using the semantics of lexical items as an indication of the copy-cognate distinction. The present study takes a different approach, using shared morphemes and their distribution as indications against code-copying. Based on the argumentation that it is difficult to account for common Japanese-Korean verb morphemes by language contact when they are also shared by the continental families, the present study aims to contribute to the reconstruction of proto-Japanese-Korean.

Fourth, by examining morphological evidence in support of the relatedness of the Transeurasian languages, the volume will complement previous studies focussing on lexical correspondences. Since the beginnings of the historical comparative study of Transeurasian languages, the emphasis has been on lexical research. Recent illustrations of this trend include the extensive collection of lexical comparisons published by Starostin, Dybo and Mudrak (2003) and my own previous work (Robbeets 2005), a collection and evaluation of existing etymological proposals relating Japanese to Korean and to the Altaic languages. It is precisel y the abundance of lexical look-alikes that has led to two diametrically opposed points of view: either that every single item is to be accounted for by code-copying or, that part of the shared vocabulary can be traced back to a common ancestor. Given the relative difficulty of copying bound verb morphology, this study can help to distinguish between the effects of contact and inheritance. This research also starts from the expectation that if a number of languages using morphological marking to express syntactic relations in the sentence are genealogically related, common ancestorship will be reflected in grammatical as well as in lexical morphemes. Moreover, demonstrating that the vocabulary and the grammar derive from the same source will enable us to reject theories of language mixing and creolization. Supporters and critics of Transeurasian relatedness seem to agree on at least this one point, i.e. that patterned morphology could substantially help unravel the question. Vovin (2005: 73) begins his critique of Starostin et al. (2003) with the postulation that “The best way … is to prove a suggested genetic relationship on the basis of paradigmatic morphology”, whereas Dybo & Starostin (2008: 125) agree that “regular paradigmatic correspondences in morphology are necessarily indicative of genetic relationship.” It is my goal to examine whether morphological evidence confirms previous findings based on lexical data.

Finally, the study is not only aimed at specialists in Northern, Central and Eastern Asian languages, but also seeks to contribute to the current literature on language classification in general. Although the search for new methods to reach beyond the limitations of the comparative method, such as Nichols’ (1992, 2003) structural comparison and Greenberg’s multilateral comparison (2000, 2002), may bring innovation and progress to the field, I aim to demonstrate that the Transeurasian languages can be shown to be related within the limits of the traditional method. Before turning to new methods, we should exhaust existing ones to the best of our abilities. It will be shown that the use of an adequate methodology allows us to move the Transeurasian family to the more established end of the classification continuuum proposed by Campbell (2003: 262; Campbell & Poser 2008: 162–163), namely from a questionable unit to a plausible and supportable family. Besides inviting historical linguistic scholarship to rethink the relatedness of the Transeurasian languages, the present study will raise a number of methodological points, especially with regard to shared grammaticalization and the distinction between code-copying and inheritance. These methodological guidelines are not without relevance for other cases of controversial classification.

The remaining sections of this introduction will define the term Transeurasian, provide internal taxonomies for the individual branches and review previous research on the topic of this book. Chapter 2 will elucidate the methods used in this work. Chapter 3 presents etymologies for verb roots including verbal adjectives and copular verbs. These are arranged according to specific phonological correspondences relevant to assessing the formal correspondences of the bound morphemes, as discussed in the following chapters. Chapter 4 to Chapter 8 propose cognate morphemes distributed over various verb categories such as negation, actionality, diathesis and tense, including nonfinite forms such as participles, verbal nouns and converbs. By way of conclusion, Chapter 9 summarizes and evaluates the correlations established in the previous chapters. It further assesses the likelihood of non-genealogical explanations and tries to find an explanation for the inconsistency of the evidence in comparison to that in support of Indo-European.

1.2 The Transeurasian languages

1.2.1 From “Altaic” to “Transeurasian”

Although Europe and Asia are physically one great landmass commonly called Eurasia, a geographical boundary between the two continents is drawn along the Ural Mountains to the Ural River and the Caspian Sea and the along the Caucasus Mountains to the Black Sea. Clearly, linguistic boundaries do not necessarily coincide with geographical boundaries. Stretching from the Pacific in the east to the Mediterranean and the Baltic in the west, the Transeurasian languages form a vast linguistic continuum that crosses the physical boundaries between Europe and Asia. Contrary to the tradition to refer to these languages as “Altaic languages,” Johanson and Robbeets (2010: 1–2) coined the term “Transeurasian” to refer to this large group of geographically adjacent languages, which share a significant number of linguistic properties and include at most 5 linguistic families: Japanic, Koreanic, Tung...

Table of contents

• Cover
• Title
• Copyright
• Dedication
• Preface
• Contents
• 1 Introduction
• 2 Methodology
• 3 Verb Roots
• 4 Negation
• 5 Verbalization and actionality
• 6 Valence and voice
• 7 Nominalization and the development of finite temporal distinctions
• 8 Converbs
• 9 Evaluation
• Endnotes
• Abbreviations
• References
• Language index
• Subject index