Grammaticalization and Language Change in Chinese

A formal view

Xiu-Zhi Zoe Wu

• Publisher: Routledge
• Year: 2004
• Print ISBN: 9780415336031
• eBook ISBN: 9781134307265
• Edition: 1
• Topic: Politics & International Relations
• Subtopic: Politics
• Index: Politics & International Relations

This innovative study on the phenomenon of 'grammaticalization' and its manifestation in Chinese provides new insights into language change in Chinese and a large number of grammatical topics. Grammaticalization occurs in all of the world's languages. Xiu-Zhi Zoe Wu demonstrates general linguistic principles present and active in the phenomenon of grammaticalization whilst also describing the modelling of language in formal theoretical approaches to syntax; so this book fills two major gaps in the current study of linguistics. Grammaticalization and Language Change in Chinese illuminates how studies of language development and change provide special insights into the understanding of current, synchronic systems of language. Using patters from Chinese, the author establishes cross-linguistic generalizations about language change and grammaticalization. This book should be of great interest to Chinese linguists and readers interested in language change in different languages.

1

INTRODUCTION

1.0 Overview

This book is an investigation of the phenomenon of grammaticalization and its manifestation in the development and synchronic status of a range of grammatical morphemes and functional categories in Chinese. There are two primary aims in the investigation: first to see how a study of various functional paradigms in Chinese may shed further light on the mechanisms of language change and grammaticalization that are generally available cross-linguistically, and secondly to examine how the use of diachronic information may help in the analysis of certain otherwise problematic synchronic phenomena in Chinese.

The book is divided into five main chapters, each of which attempts to address the above-stated dual aims of: (a) providing a synchronic analysis of a certain functional paradigms in Chinese assisted by considerations of how functional categories may develop over time, and (b) investigating the nature of general processes of grammaticalization. Concerning the latter goal, each chapter will argue for and attempt to illustrate a different process of grammaticalization at work in Chinese, and will show how grammaticalization may have important effects on aspects of surface linear word-order, frequently causing significant distortions of the underlying syntactic structure.

General conclusions of the investigation are provided chapter by chapter and also outlined below in a brief preview of the various phenomena to be considered. Prior to this however, the chapter provides a short background introduction to certain general ideas relating to grammaticalization and explains the rationale for applying a formal theoretical approach to the study of grammaticalization phenomena in Chinese. Concerning the technical framework and theoretical assumptions adopted in the book, analysis is carried out in a broad Minimalist/Chomskyean approach which assumes a transformational component (i.e., movement) and two interface levels, PF and LF. As in Chomsky (1993, 1995), the combination of lexical items via Merge into initial syntactic structures is assumed to lead first to a point of Spell-Out feeding PF and phonetic interpretation, and then derivationally continue on to a level of LF and semantic interpretation. Various other assumptions relating more specifically to language change and the phenomenon of grammaticalization will be introduced as the chapters proceed and as such notions become particularly relevant.more specifically to language change and the phenomenon of grammaticalization will be introduced as the chapters proceed and as such notions become particularly relevant.

1.1 Background: Grammaticalization as an Area of Study

Grammaticalization as a term was first used by the French linguist Antione Meillet in a paper written in 1912. Commenting on Meillet’s paper, Ramat (1998; p. 108) observes that gramaticalization is assumed to have two rather different functions:

(a) to create new forms that replace old forms in existing grammatical structures, which remain essentially the same from the point of view of function,

(b) to introduce into grammar new categories, i.e., new units of form/function.

Most commonly, grammaticalization is assumed to involve some kind of reanalysis, involving the development of a word/morpheme into a grammatical marker of some type. In many cases this may involve the development of a word which has descriptive, referential content (e.g., a noun, adjective, verb, etc.) into a grammatical morpheme/word which has a predominantly functional role (e.g., a complementizer, tense-mood marker, determiner, etc.), as schematized in (1). The term “grammaticalization” here is sometimes used to refer to the results of such a categorical conversion (e.g., the creation of a new tense-marker from a descriptive verb), and also to the mechanisms and processes which actually give rise to the reanalysis.

(1) Common shift in category type in grammaticalization lexical word → grammatical word/morpheme (e.g., verb → complementizer, noun → agreement-marker)

In all occurrences of grammaticalization there is some change in the meaning of the grammaticalizing element. Characteristically this is described as involving changes from more referential meanings to increasingly more abstract meanings. Hence, for example, a word with clear descriptive, “referential” content such as (a person/animal’s) ‘back’ denoting the tangible part of a physical body might become re-employed as a more abstract temporal relational term with the meaning ‘X-ago (X = point in time)’, and syntactically come to function/grammaticalize as an adposition in some languages.

(2) Typical shift in meaning accompanying grammaticalization: more descriptive/referential → more abstract

Much of the work that has been carried out in the study of grammaticalization in the last century has focused on trying to understand and explain how the shifts of meaning which occur in instances of grammaticalization can be viewed as natural paths of evolution, with a range of cognitive and pragmatic explanations being given for the variety of changes attested (e.g., Heine, Claudi and Hiinnemeyer 1991). Such work frequently describes changes in terms of “grammaticalization chains”, in which one link (stage) in a path of development naturally leads to a following link/stage via processes of pragmatic enrichment and inferencing. Expansion and development of meaning here is sometimes suggested to be due to the cognitive mechanism of (permitting) metaphor, as for example in the common creation of future auxiliary verbs/expressions from verbs denoting motion to some distant point (English ‘going to’: I am going to leave, French aller. Je vais partir), the mental image of motion towards an entity allowing for itself to be re-interpreted metaphorically as ‘motion’ of the subject towards a future point in time. Growth in the strength of pragmatic implicatures made when linguistic expressions are used has similarly been noted to be a cause of meaning change which can over time result in grammaticalization, with secondary inferred meanings coming to be more and more automatically associated with the use of a word/expression (see in particular Konig 1988, Traugott and Konig 1991). Two examples of abstract grammaticalization chains are given in (3) and (4) from Bybee, Perkins and Pagliuca (1994, chapter 7). These represent common routes of development in which verbs with the meanings given in small capitals on the left have been noted to frequently develop/grammaticalize into verbs with meanings on their right. Hence in (3), a verb encoding the meaning of DESIRE (e.g., ‘want’) may develop the meaning of WILLINGNESS, and then INTENTION before finally becoming used as a verb of prediction/a future tense verb.

(3) DESIRE > WILLINGNESS > INTENTION > PREDICTION

(4) MENTAL/PHYSICAL ABILITY > ROOT POSSIBILITY > EPISTEMIC POSSIBILITY

In addition to comprising a series of two or more links corresponding to different but relatable meanings, grammaticalization chains are regularly taken to have three further properties. The first of these is that the process of grammaticalization frequently results in increased phonetic reduction and phonological dependency on some other host word. It has been observed that independent descriptive words may first develop into syntactically independent grammatical words, and then with increased use often undergo phonetic reduction and begin to display semi-dependent clitic-like status. Finally such elements may become fully dependent on a restricted lexical host and so evolve into bound morphology. Such a chain of development is indicated in (5), and reflects Givóri’s (1979) oft-quoted statement concerning the potential effects of grammaticalization: Today’s morphology is yesterday’s syntax’.

(5) Route of development for elements grammaticalizing as bound morphemes:

lexical word → grammatical word → grammatical clitic → bound morpheme

A second property commonly assumed to constrain the development of one meaning of a word/morpheme in a grammaticalization chain into a second, different meaning is the occurrence of periods of overlapping variation. If a certain word/morpheme has an original meaning A, and this over time develops into a different meaning B, it is argued that there will always be some intermediate period where the relevant word/morpheme can potentially encode either the meaning A or the meaning B (Hopper and Traugott 1993). In such a period of variation before the (frequent) disappearance of the original meaning A, there will hence be a stage in the grammaticalization chain in which a single morpheme is ambiguous between the newly emerging meaning and the earlier original meaning. In this sense, grammaticalization is often taken to be a gradual process of change with no sudden and abrupt changes in meaning:

(6) A → A/B → B

The third property argued to characterize grammaticalization chains with very limited exception is the property of unidirectionality of development. This is a claim, resulting from much observation of patterns of grammaticalization, that any development of a word/morpheme down the path of a grammaticalization chain is irreversible, and there can be no undoing of any change and regression of a morpheme to an earlier stage in the grammaticalization chain once a change has been properly effected. The only way for a word/morpheme to develop over time is therefore always forwards to new stages in a grammaticalization chain, and not back to earlier/prior stages. This constraint is assumed to affect both the development of meaning from more referential to less referential (2), and the progression of increasing phonological dependency schematized in (5). Consequently, with regard to meaning, if a morpheme following the path in (3) has developed the meaning of INTENTION and fully lost the earlier interpretation of DESIRE, it is argued that it will never re-develop the interpretation of DESIRE (or WILLINGNESS) from a loss of the meaning INTENTION. Rather, the only way for such a morpheme to develop is to a new, less subject-centered meaning such as PREDICTION. Similarly, concerning the developmental pattern indicated in (5), if an element has developed into a bound morpheme, it is assumed that it will never later redevelop its phonological independence and revert to one of the earlier (leftward) stages in the sequence of development.

Summarizing briefly now, most work which has been (and still is) carried out on grammaticalization phenomena is primarily concerned with (a) charting the development of descriptive words into increasingly more grammatical forms, and (b) attempting to find plausible cognitive and pragmatic explanations for the changes of meaning which occur during the course of grammaticalization. Such work has led to the important discovery of a whole range of developmental patterns and the suggestion that grammaticalization often conforms to many typical routes of development referred to as grammaticalization chains.

1.2 A Formal Approach to Grammaticalization

Although the majority of research into grammaticalization has been developed within functionalist and cognitive grammar-type frameworks, the current study of Chinese makes use of a formal theoretical framework to approach grammaticalization phenomena, and specifically the set of theoretical assumptions made in recent Chomskyean-Minimalist approaches to language (in particular those in Chomsky 1993, 1995). There are a number of reasons for selecting a formalist approach here, and I outline the rationale for this below.

One important reason for making use of a formal syntactic framework for the present study relates to the na...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Acknowledgments
• 1. Introduction
• 2. The Classifier GE: Movement And Reanalysis
• 3 Relative Clause DE: Directionality, Clausal Raising And Sentence-Final Particles
• 4. DE In Focus Sentences: From D To T
• 5. Resultative Constructions: Directionality And Reanalysis
• 6. Verbal LE : Aspect And Tense
• 7. Post-Word
• Bibliography
• Index