Interactive Morphonology

Metaphony in Italy

Martin Maiden

• Publisher: Taylor & Francis
• Year: 2018
• eBook ISBN: 9780429888274
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

First published in 1991. The existence of morphonology had been the subject of intense debate in twentieth-century linguistic theory. Attempts to identify putatively morphonological phenomena had often foundered on the widespread assumption of a rigid dichotomy between synchronic morphological structures and the phonetic processes which historically shared them. With the difficulties of establishing any role for morphonology clearly identified, the author introduces a comparative and historical survey of the morphologization of metaphony in Italian dialects. On the basis of this the existence is argued of authentic synchronic 'morphonological' interaction between morphological structures and phonetic processes, such that inflectional paradigms serve to specify phonetic details of implementation of incipient sound changes. The circumstances under which such interaction may be expected to occur are discussed.

This book is an important contribution to our understanding of both morphology and phonology, taking seriously the implications of abandoning a rigid distinction between synchronic morphology and diachronic phonology. It successfully integrates linguistic theory with the analysis of philological data, and indicates the direction for future research on morphonology. This detailed study of Italian dialects also constitutes a valuable addition to the study of Romance dialectology.

1

INTRODUCTION

1.1 The problem

This book is an investigation of the relationship between phonetic processes and morphological structure. The type of question with which it is concerned may be illustrated on the basis of the following fragments of the verbal paradigm in three Italian dialects (A, B and C), illustrated in Table 1:

Table 1

A	B	C
korro	korro	korrə	‘I run’
korri	kurri	Kurrə	‘you run’
korre	korre	Korrə	‘he runs’

In the fragment from dialect A, the morphosyntactic category of person is signalled solely by a set of inflectional suffixes. In B, in addition to the inflectional suffixes, there is an alternation in the stem vowel arising from METAPHONY, a phonetic assimilatory raising of stressed vowels, conditioned by a following unstressed high non-mid vowel (/i/ or /u/). As for C, a phonetic centralization process has neutralized the distinction between unstressed vowels, obscuring the original conditioning environment for metaphony: the metaphonic alternation remains as the sole indicator of the morphosyntactic category of person, originally signalled by the inflection. In other words, metaphony is clearly MORPHOLOGIZED. The nature of morphologization will be a central theme of this study, but it is very difficult to extract from the linguistic literature a coherent initial consensus on the phenomenon (cf. Roberge (1980:2f.)). The following, extremely general, formulation (similar to that offered by Kenstowicz and Kisseberth (1977:74)) might be taken as a starting point:¹

Morphologization occurs when some aspect of a phonetic change is associated by speakers with some morphological category.

But one is bound to ask what ‘aspect’ of the phonetic change is involved, and under what circumstances such ‘association’ occurs. The answer may seem, in part, obvious: given that in dialect C the phonetic conditioning environment for metaphony has been altered and has merged irrecoverably with the reflexes of other unstressed final vowels, which never triggered metaphony, speakers of C must perforce associate the alternation, resulting historically from metaphony, with the morphosyntactic category of person. This may seem an unexceptionable claim, but we shall see that it is potentially controversial: some scholars would argue that the very existence of morphological paradigmatic alternation should not be assumed a priori. As for dialect B, the scope for controversy is more evident. Simple observation of the synchronic data permits a range of initial hypotheses about how speakers might account for metaphonic alternation. These might be purely phonological (as in (a) below), or purely morphological (as in (b)), or alternatively might combine, apparently redundantly, both morphological and phonological conditioning (as in (c)):

a. Phonological

i. Metaphonic alternation is produced by the phonetic assimilatory process.

ii. Metaphonic alternation is phonological, but not phonetic. The original phonetic motivation has disappeared, and speakers have a more or less arbitrary phonological rule substituting stressed /u/ for /o/ in the environment of a following unstressed /i/ or /u/.

b. Morphological

i. The metaphonic alternant is associated with -i, the inflectional exponent of [second person].

ii. The metaphonic alternant is associated directly with the morphosyntactic property of [second person].

c. Phonological and morphological (combinations of (a) and (b) above)

i. (ai) + (bi)

ii. (ai) + (bii)

iii. (aii) + (bi)

iv. (aii) + (bii).

Much of this book will be dedicated to establishing which of these analyses speakers have really made. Evidence will emerge in favour of (ci) or (cii): phonetic and morphological conditioning coexist; also, metaphonic alternation is probably directly associated with the morphosyntactic category, independently of the inflectional suffix. I shall show, moreover, that this apparently redundant combination of a phonetic process and morphological conditioning is not redundant at all, but involves an authentic interaction of phonetic and morphological factors, having distinctive and unique characteristics of its own.

1.2 The question of morphonology

The question of the relationship between phonetic processes and morphological paradigms has frequently occupied linguists over the past century. An account of theoretical approaches to this problem will be given in the following chapters. Suffice it to say here that, until recently, there had been a strong tendency to separate phonetic and morphological phenomena. In the late nineteenth and early twentieth centuries (cf. the writings of Kruszewski, Baudouin de Courtenay, Saussure) such separation followed from the lucid perception of the distinction between linguistic synchrony and diachrony, and in part reflected a reaction against the predominantly diachronic and principally phonetic preoccupations of Neogrammarian practice. Part of this reaction was the assumption of a dichotomy between alternation produced by ‘live’ phonetic processes, and alternation which was synchronically associated with morphological categories, and whose phonetic motivation was purely a matter of diachrony. An early, and characteristically farsighted, challenge to this dichotomy was issued by Sapir (1921:183f.):

I am inclined to believe that our present tendency to isolate phonetics and grammar as mutually irrelevant linguistic provinces is unfortunate. There are likely to be fundamental relations between them and their respective histories that we do not yet fully grasp.

The study of morphophonology,² a term first given wide currency (in the significantly haplological form morphonology) by Trubetzkoy in 1929, originally dealt with the phonological shape of morphemes in general. Its main concern, however, came to be the description of the nature and function of phonological oppositions within morphological paradigms, and of the phonological environments in which such morphologically correlated oppositions were neutralized. In a forceful critique of classical morphonology, Martinet (1965) challenged the legitimacy of any postulated independent ‘morphonological’ domain: apparent alternations whose distribution is statable exclusively in terms of their phonological environment should be assigned to ‘phonology’, while morphological alternations not uniquely attributable to phonological environment should be relegated to ‘morphology’. At the root of this distinction was a perception of the synchronic arbitrariness of the latter type of alternation: detached from its original phonological environment, ‘morphonological’ alternation merely becomes an arbitrary taxonomy of those phonemic oppositions whose distribution happens to coincide with alternating morphosyntactic categories. Indeed, such cases might not be essentially different from, say, paradigmatically alternating suffixes (e.g., the verbal inflections -o, -i, -e in the Italian dialects), each of which may have historical origins wholly independent from those of the other inflections.

One certainly finds a wide consensus among linguists that there exists no independent domain of morphonology (cf. Kilbury (1976:56–68)). Dressler (1985a) surveys the gamut of possible criteria for the establishment of such an independent domain, and concludes that the range of putative ‘morphonological’ phenomena can in general be exhaustively factored into either phonological or morphological elements. However, the past twenty years have seen two notable interpretations of morphonology as a kind of ‘interface’ or interaction between morphology and phonology. One of these is Matthews (1972b) who, reacting against Martinet, argues for a conception of morphonology in which morphologically specified alternants furnish a ‘resolution’ of ‘live’ phonological constraints. The second is Dressler’s study, alluded to above, which conceives morphonology as a domain of synchronic interaction between phonological and morphological rules: in Dressler’s approach, a rule is morphonological to the extent that it simultaneously possesses some degree of morphological and phonological ‘naturalness’, the latter being usually identifiable with phonetic motivation. Dressler provides a wide (if not uniformly convincing) range of diachronic and acquisitional substantive evidence for the psychological reality of ‘interactive’ morphonology, which suggests that morphological rules having a high degree of phonological ‘naturalness’ are favoured in language retention and acquisition.

What I offer in this volume is a new kind of substantive diachronic evidence for an ‘interactive’ conception of morphonology. I assert, in effect, that there exist phenomena which are simultaneously, and non-redundantly, both phonetically and morphologically conditioned. This view owes part of its inspiration to the studies by Matthews and Dressler cited above even if, as will become apparent, I dissent in some measure from their interpretation of the data they present. Most of Dressler’s examples involve rules which are morphonological in that they apparently retain some element of their original phonetic causation, but such examples are potentially open to the challenge that persistence of apparent ‘naturalness’ is really synchronically redundant or arbitrary (cf. Hellberg (1978)). Moreover, morphologization is generally assumed by Dressler to entail some decrease in the phonological ‘naturalness’ of a rule. I document the emergence of a variety of Italian metaphony whose naturalness (in the sense of phonetic motivation) is, on the contrary, as great in certain morphologically specified environments as anywhere else in the grammar. Such a rule simply cannot be factored into redundant, simultaneously operating, phonological and morphological conditionings. Seeking to identify the circumstances in which such interaction occurs, I reach a position akin to Matthews’: the operation and resolution of incipient phon...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Acknowledgements
• Abbreviations
• 1 Introduction
• 2 Phonetic and morphological conditioning
• 3 Morphonology and morphologization
• 4 Morphological conditioning of phonetic processes?
• 5 The interactive view of morphonology
• 6 Origins and nature of Italian metaphony
• 7 Metaphony in the inflectional paradigms
• 8 The metaphonic evidence
• 9 Phonetic vulnerability in diffuse paradigms
• 10 Characteristics of morphonology
• Appendix: Italian and Swiss localities cited
• Map 1 Italy: regions and provincial centres cited
• Bibliographical references
• Index