Mutation in Welsh

Martin J. Ball, Nicole Müller

• Publisher: Taylor & Francis
• Year: 2002
• eBook ISBN: 9781134969326
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

The Initial Consonant Mutation system of Welsh is unique to Indo-European languages and has been the subject of much theoretical research. The multi-faceted nature of the phenomenon demands multi-dimensional treatment and this uniquely comprehensive book provides an integrated overview of this important feature from a wide linguistic viewpoint. In Welsh, Initial Consonant Mutation has implications for historical and comparative analyses, phonetic description, phonological theory, syntactic theory, and the interfaces between phonetics and phonology, morphology and phonology, and phonology and syntax. It also requires examination from semantic, psycholinguistic and sociolinguistic perspectives. This study, therefore, brings together a variety of approaches to a wide range of levels of linguistic analysis, all concentrated on one unusual linguistic feature. A detailed review of past research, together with an exploration of recent theoretical advances in many areas, makes this an indispensable book for departments of Celtic Studies and all scholars of comparative linguistics.

1 Mutations in Standard Welsh

1.0 INTRODUCTION

We will start our investigation of the initial consonant mutations in Welsh by examining how this phenomenon operates in the modern standard language. Before we can go into detail on this, however, we must first define more strictly than was done in the Introduction what we mean by mutation, and also what we mean by Standard Welsh.

This discussion will be followed by an account of the three consonantmutation types, and the syntactic triggers of these in the modern standard language. Also the related phenomenon of pre-vocalic aspiration will be described (pre-vocalic nasal prefixing as in Irish a n-athair, ‘their father’, is clearly a similar process, but does not occur in Welsh).

1.1 INITIAL CONSONANT MUTATION

As noted in the Introduction, this process is on the borderline of phonology and syntax, in that the changes that occur are phonological while the circumstances that produce the changes are not themselves phonological but, in a broad sense, syntactic. This means that initial consonant mutation is unlike most other phonological processes recorded in natural languages, where the changes are triggered by a particular phonological environment or by a particular suprasegmental setting (e.g. allegro speech).

1.1.1 Definitions

Phonological changes triggered by phonological environments are, of course, relatively straightforward to capture in phonological rule formation, but initial consonant mutation (ICM) presents problems, as the environment triggering the change is syntactic. Nevertheless, as a starting point, the following preliminary definition and rule format is proposed.

(1.1) ICM is a process whereby word-initial consonants undergo one of three sets of phonological changes when in certain syntactic environments. The particular set of changes is determined by the syntactic environment.

(1.2) ICM: ##CX → ##C’X/SE (where ## stands for word boundary, C and C’ represent consonants related via a phonological process, X represents any string of consonants and vowels allowed to follow C by phonotactic constraints, and SE stands for syntactic environment.)

This particular formulation will be returned to, and modified in the light of our discussions, in later chapters.

Clearly, both the types of phonological change and of syntactic environment need to be specified further, and these will be returned to in 1.3–1.6 below. It should be noted here that only mutation processes affecting word-initial consonants are discussed in detail in this book; similar patterns morpheme-initial but word-internal do occur, and traditional grammars often include these as part of their discussion on mutations. We will look at word-internal consonant changes in connection with the historical development of ICM in Chapter 3. These word-internal changes are, of course, part of the derivational morphology of the language, and are looked at briefly in Appendix 1. However, they do represent the main consonant-mutation types described below (plus one other: ‘hard mutation’, see Appendix 2), and they are of interest in that, as with the syntax, specific mutation triggers (in this case affixes) cause specific mutations. Unlike word-initial mutations, however, the process is not dynamic: the word-internal mutations are static and do not have to be ‘constructed’ by the speaker as word-initial changes have to be (if we assume that derivationally complex words are stored as whole units rather than as parts; see discussion in Chapter 9 on lexical storage in Welsh).

1.1.2 Terminology

In the following discussions of mutations throughout this book we employ a set of terminology based on a classification of this feature described below. The mutation types themselves are termed Soft Mutation (SM), Nasal Mutation (NM), Aspirate Mutation (AM) and Pre-Vocalic Aspiration (PVA). The non-mutated form of a consonant is termed the radical. Descriptions of the processes involved and alternative labels are given in 1.3–1.6.

Firstly, the syntactic environment (SE) of a mutation is characterized as consisting of a trigger and a target. The trigger is whatever feature (see below) which, synchronically at least, can be thought of as ‘causing’ the mutation. The target is the word ‘receiving’ (i.e. undergoing) the mutation (Tallerman’s 1987 usage of mutatee for target is rejected due to its artificiality). Whereas usually the relation between these two is relatively free (i.e. the target can be any word that fits the word-class of the target), there are occasionally restrictions holding between the trigger and target (see, for example, in section 1.4 below).

Secondly, the triggers themselves have often been classified into types and sub-types (see accounts of previous classifications in Chapter 2). The most common trigger type we term lexical trigger, which covers instances where a single lexical item can be identified as the source of the mutation. The pure lexical trigger consists of a specific lexical item that always ‘causes’ a particular mutation to the following lexical item, the only restriction being that the target has to be of the word-class(es) able syntactically to follow the trigger (semantic restrictions may also play a part, but only so as to produce semantically well-formed utterances). For example, ei gath (‘his cat’ ei+SM); ei chath (‘her cat’ ei+AM); fy nghath (‘my cat’ fy+NM); radical: cath ‘cat’. In these examples ei ‘his’ is seen as a pure lexical trigger of SM, ei ‘her’ as a pure lexical trigger of AM, and fy ‘my’ as a pure lexical trigger of NM.

The categorial lexical trigger is a less common variety of this type. Here one category of a potential lexical trigger (rather than a whole class or a specific item) is crucial to the mutation process. For example, the abbreviated pronoun ’i ‘him, his’ only causes PVA when in the accusative (object) case, but not when in the genitive (possessive) case; for example, fe’i hachubwyd ‘he was saved’ as opposed to a’i arian ‘and his money’. The abbreviated pronoun ’i ‘her, hers’ does cause PVA in the genitive, however. An alternative solution could clearly be devised in this instance whereby ’i is established as the homophonous realization of four different pronominal forms; indeed, some of the other categorial triggers listed for the different mutations below could be accounted for in a similar way. We retain the category, however, to account for those instances which cannot so be avoided, and to leave open the possibility of monolexical analyses of forms such as the pronouns noted above.

This type should not be confused with, for example, the mutation of feminine nouns following the definite article (though see Chapter 6 for an alternative account of this context). Here, the target is the item which is restricted, while the trigger is not. Such examples we term restricted lexical triggers, and include instances where a trigger only causes mutation on individual lexical items. Finally, the debate as to whether some or all of the lexical triggers noted below are in fact separate words, rather than clitics, is left to Chapter 7, where notions of phonological words as opposed to morphological words are dealt with.

The remaining main trigger type is the syntactic trigger, which covers those instances where no single lexical item can be identified as the source of the mutation; rather, the syntactic/semantic frame (perhaps better expressed as a ‘functional frame’) itself is the trigger. This category can be considered as having two sub-types: the pure syntactic trigger and the restricted syntactic trigger. An example of the former is the SM of object NPs of inflected verbs, this trigger being subject to no further conditions; while the latter is exemplified by SM in apposition, which is restricted lexically, in that with personal names, for example, some combinations do take the mutation while others do not.¹ We use the term restricted syntactic trigger to cover restrictions due to lexis, syntax or semantics, or register.

Some apparent examples of syntactic triggers result from the so-called ‘zero lexical trigger’; that is, the lexical trigger has been subject to an optional deletion rule dependent on register. Although other treatments have proposed a special category of mutations with no trigger, we class these as varieties of the pure lexical trigger, and examples are given in sections 1.3 and 1.5 below. We return to...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Tables
• Preface
• Acknowledgements
• Abbreviations
• Introduction
• 1 Mutations in Standard Welsh
• 2 Previous accounts of mutation
• 3 Comparative and historical aspects of mutation
• 4 The phonetics of mutation
• 5 The phonology of mutation
• 6 The syntax of mutation
• 7 Interfaces
• 8 The semantics of mutation
• 9 The psycholinguistics of mutation
• 10 The sociolinguistics of mutation
• 11 The future of mutations
• Appendix 1: The morphology of mutation
• Appendix 2: Hard mutation
• Appendix 3: The pronunciation of Welsh
• Notes
• References