Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 12 November 2019


Abstract and Keywords

This chapter introduces the key elements of inflection, the expression of grammatical information through changes in word forms, and how it differs from derivation. It then outlines the structure and content of the volume and discusses the main topics covered: the fundamental building blocks of inflectional form and content in part one; paradigmatic structure and mapping between function and form in part two; change and variation over time in part three; computational issues from a theoretical and practical standpoint in part four; psycholinguistic considerations in part five; and studies of specific inflectional systems in part six .

Keywords: inflection, inflectional form, paradigmatic structure, computational issues, psycholinguistics

1.1 Inflection

Inflection is the expression of grammatical information through changes in word forms. For example, in an English sentence such as my cousin admires you, although the subject is clear from its placement before the verb, it is also indicated by the final -s on the verb, which tells us that the subject is third person singular, and hence cousin and not you, and also that the verb is in the present tense. This is fairly straightforward: we have an easily indefinable element (suffix -s) with an easily identifiable function (marking the presence of a third person singular present tense subject), but it seems to be in the nature of words that they may well follow their own rules, and that these rules may get quite a bit more involved. For example, the selection of forms in (1) from Chiquihuitlán Mazatec, an Otomanguean language of Mexico, shows how difficult it may be to say where exactly the grammatical information is being expressed. Subject marking is distributed across the word form, involving the initial segments, for example hu- versus čho-, the final vowel, e.g. -æ versus -e, and tone, for example high (marked here as superscript ‘1’) versus mid-high (superscript ‘2’). Further, each word in this little sample has a different pattern of alternation for all three of these elements.

  1. (1) Introduction

As impressive as such systems may be, it is perhaps an even more striking fact that, from the point of view of language as a whole, inflection is apparently superfluous. While every language operates by stringing together words, manipulating the shapes (p. 2) of words is further elaboration that many are happy to do without. This is probably what has lent inflection the curious and often contentious status it has within linguistics. On one view, inflection is just a particular instance of the more general lexical, semantic, syntactic, and phonological properties needed to characterize language as a whole, so that a word form such as admire+s is the syntactic combination of lexical units in the same way that the sequence my + cousin + admires is. Word-internal syntactic and phonological rules may differ from those that apply to other domains, but they are all cut from the same cloth. However, others see the behaviour of inflected forms— both the relationships that obtain between the forms of a single word, and those that obtain across different words—as potentially too divergent from the general set of linguistic modules for there to be a reliable mapping between them, and so postulate a distinct morphological component to grammar, subject to its own rules.1 Obviously, taking one or the other approach presupposes different models of the architecture of grammar, which in turn encourages different lines of enquiry. In the absence of an autonomous morphological component, questions of morphophonology and feature structure may assume prominence, whereas the assumption of an autonomous morphological component may foster a greater focus on questions of paradigm structure. But however they are interpreted, there is a fascinating and well-delimited set of facts out there to be explored. And if inflection is not shared by every language, it is itself a uniquely linguistic phenomenon, without obvious parallel in other communicative and symbolic systems.

Inflection shares the larger realm of morphology with derivation, with which it has much in common, so it is important to distinguish the two in order to delimit the scope of the present volume. The basic difference of course is that inflection produces different forms of a word, while derivation produces new words. This is a rather slippery definition, as it presupposes that we can distinguish between different forms of a word and different words, but at some level we must all recognize that while the word cat is related to both cats and kitten, the nature of the relationship is different, both in terms of meaning and in terms of form. More precise attempts at a definition are of necessity multi-pronged. For example, Stump (1998: 15–18) offers the five diagnostics in Table 1.1. The first is in a way the most obvious: inflection preserves the lexical meaning while derivation changes it. The other diagnostics are essentially definitions of inflection, with derivation defined negatively: inflection involves a predetermined set of meanings (morphosyntactic values) which a word is required to accommodate, with the term derivation then applying to a looser constellation of morphological operations.

The key point is that making a distinction between inflection and derivation leads to different expectations. With inflection, the set of word forms is predetermined, in that (p. 3) a member of a given word class2 has certain duties to perform. For example, a verb X in English needs to be available for use in such sentences as She often X1 or You X2 yesterday. Where X = ski the relationship between X1 and X2 is transparent (ski-s, ski-ed), but the expectation that we should be able to put any verb into these contexts forces us to look beyond obvious formal resemblances, sanctioning such strange bedfellows as go and went. Derivation for its part has no preset list of job duties, and thus no binding obligations. The fact that we can derive duckling from duck does not sanction the expectation that there should be a comparable derivative of turkey (not for this author, at any rate), and a paraphrase or compound is good enough (say, turkey baby or turkey chick). As a consequence, the notion of derivation is in practice reserved for those instances where there is a formal resemblance between items that can be expressed as a morphological rule. Thus while the semantic relationship between puppy and dog is the same as that between duckling and duck, few would want to label this a derivation, and the status of gosling vis-`a-vis goose (in any synchronically meaningful sense) will depend in part on how much faith one puts in the morphological rules needed to relate them.

Table 1.1 Inflection versus derivation, per Stump (1998)



Lexical meaning and/or part of speech









Semantically regular



Closure (recursiveness)



Thus for any given language, inflection embraces a more-or-less closed system of interrelated functions and forms. In a sense, we can treat inflection as a discrete and self-contained slice of language as a whole, in the same way that we can treat language itself as a discrete slice of human cognition. This does not necessarily mean committing oneself to the notion that there is a component of language exclusively dedicated to inflection (or morphology more generally), any more than talking about language presupposes belief in an autonomous language faculty. And whatever status we accord an interface phenomenon such as morphosyntax—whether as syntax, morphology, or a domain of its own—it is evident that the existence of inflected word forms has contributed materially to the terms that we use to talk about it, and the concepts behind it.

The following chapters are meant to address inflectional morphology from as many thematic angles as possible. Particular attention has been paid to taking examples from a wide variety of languages. This helps ensure that the topics covered are of broad (p. 4) cross-linguistic relevance and makes for a lively and varied presentation. There are no chapters explicitly devoted to the exposition of particular morphological theories. This is intentional, and comes from the conviction that the most important difference between theories is the phenomena they choose to focus on, not the technicalities of their implementation. The approach taken here is therefore ecumenical, and the chapters come from various theoretical perspectives, as appropriate for the topic. We hope that this will be a useful and engaging resource for the study of what we find one of the most fascinating aspects of language.

1.2 About the Volume

The first part covers the fundamental building blocks of inflectional systems. The morpheme (Chapter 2) has long served as the basic unit of morphological analysis, an almost unavoidable point of reference even within those schools of thought that reject it on principle. Anderson traces the history and differing conceptualizations of the notion, showing far greater divergence than is usually acknowledged. Features (Chapter 3) characterize the content side of inflection. Rather than treat each inflected form as expressing an isolated unit of meaning, we typically see them as manifesting particular values of higher level and possibly intersecting categories, such as case and number. The logic of feature systems is sometimes at variance with their formal expression, leading to some descriptive and theoretical challenges. Exponence (Chapter 4) describes how inflectional feature values are translated into morphological forms. A common assumption is that we can distinguish concatenative morphology, which involves the addition of discrete phonological units, from non-concatenative, which involves transformational operations or mapping onto preset templates. This chapter takes a more nuanced approach, using a finer-grained classification to integrate both types.

The second part focuses on what is probably the most characteristic property of inflectional systems, paradigmatic structure: its nature, its variants, and its interfaces with phonology and syntax. The general concept of the paradigm is the topic of Chapter 5, which surveys the various ways it has been construed in different morphological traditions. Paradigms have assumed particular relevance in recent work which examines the implicational relationships that obtain between inflected forms, using insights offered by the application of information theory. The systematic variation between functionally equivalent but formally distinct alternative paradigms that occurs in many languages results in inflection classes. Chapter 6 addresses the theoretical, typological, and diachronic issues that they raise, considering the ways in which they may best be represented and accounted for. Inflection classes represent one dimension in which the mapping between function and form is not straightforward, since they seem to involve arbitrarily different patterns of inflectional exponence across (p. 5) different lexemes. Cross-cutting this we also find deviations within the paradigm, surveyed in Chapter 7, manifested by such phenomena as syncretism, deponency, and defectiveness, each involving a less-than-straightforward relationship between morphosyntactic values and the forms that realize them. While this might be taken as evidence of purely morphological forces determining the structure of the paradigm, an alternative approach is to attribute some measure of these paradigmatic effects to phonology, as explored in Chapter 8. Chapter 9 steps beyond the strict confines of the inflected word form to look at periphrasis, where the inflectional paradigm appears to embrace multi-word constructions, including borderline cases involving clitics, whose status lies somewhere in between that of a free word and an affix.

The third part covers change and variation over time. Chapter 10 surveys fundamental issues in diachrony, offering a typology of morphological and morphosyntactic change, as well as providing a picture of the larger context, including cycles of change and the role of other linguistic components as a triggering factor. Since inflectional morphology is generally resistant to borrowing, the topic of contact-induced change is relatively understudied. Chapter 11 presents the borrowing of inflectional morphology as the interplay between two factors: the scope of borrowing (inflected word forms or isolated inflectional material) and its influence on morphosyntactic categories already existing in the language.

The fourth part covers computation, both as a tool for theoretical exploration and for practical applications. Chapter 12 looks at modelling inflectional structure computationally, as a way of clarifying the nature of inflectional operations and comparing alternative treatments. While these techniques involve explicit writing of inflectional rules in computationally tractable terms, machine learning (Chapter 13) can be used to explore more abstract questions about the nature of rules, the trade-off between rules and listing, and the role of linguistic universals versus learning biases in the development of inflectional structure. Although techniques of machine learning can be construed as modelling at least some aspects of human language acquisition, machine translation (Chapter 14) understandably takes a pragmatic approach to the issue, with inflectional morphology posing a particularly difficult challenge. In this quickly developing field we see an increasing convergence between rule-based and statistical methods, an interesting parallel to what we also see in purely theoretical treatments of inflection.

Part V approaches the relationship of inflection to the human mind from two perspectives. Chapter 15 looks at the acquisition of inflectional morphology, with a particular focus on cross-linguistic variability. It argues for a step-by-step process of schema acquisition and extension, rather than a dual-route split between storage and rules. Chapter 16 explores the effects on inflection of a range of neurocognitive disorders. In contrast to the preceding chapter, it is suggested here that the evidence is consistent with a dual-route model of inflection, with damage to semantic memory affecting irregular (i.e. stored) morphology, and disorders affecting frontal/basal ganglia (related to motor functions) affecting regular (i.e. rule generated) morphology.

(p. 6) Part VI is devoted to sketches of individual inflectional systems, illustrating a range of typological possibilities across a genetically diverse set of languages. Iha (Chapter 17) is a Trans New Guinea language spoken on the western tip of the island of New Guinea, and is particularly notable for the range of argument categories marked by agreement on the verb, with grammatically aligned suffixes (marking subject and object) interacting with semantically aligned prefixes (patient versus other). Pulaar (Chapter 18) is the westernmost dialect of Fula (Atlantic, Niger-Congo), spoken in Senegal, and is known for its heavy reliance on consonant mutation as an inflectional exponent. As with many other Niger-Congo languages, there is an extensive noun class system, and corresponding to this a large gender system. The conflicts in gender that would arise in conjoined noun phrases cause real difficulties, and are usually avoided. A striking feature of its morphosyntax is the relative tense, which involves a peculiar constellation of morphological and syntactic properties (e.g. inversion of verb–pronoun order), in a set of constructions which so far have defied precise characterization. Lithuanian (Chapter 19) is a Baltic language of the Indo-European family, known for a number of conservative features, including its complex nominal paradigms, involving both suffixal and prosodic alternations. A number of points of special interest are discussed here, including (i) the quasi-agglutinative local cases, calqued from the corresponding constructions in Finnic, (ii) the doubly problematic status of the reflexive, which is a borderline case between inflection and derivation on the one hand, and between clitic and affixal status on the other, (iii) evidentials, an unusual feature for an Indo-European language, and (iv) aspectual distinctions, which are only weakly grammaticalized and hence of uncertain status. Chamorro (Chapter 20), an Austronesian language of the Marianas islands (including Guam), stands out among the languages illustrated here in the categorial indeterminacy of its roots, which can be inflected as nouns or verbs, its ergative alignment, and the use of infixation and reduplication as productive means of inflectional exponence. Verbs in Murrinh-Patha (Chapter 21), a non-Pama Nyungan language of northern Australia, display a complex two-part structure, where lexical stems combine with members of a small set of highly irregular classifier stems, so that grammatical and lexical meaning are distributed across these two word forms. The resulting system is rife with discontinuous dependencies, multiple exponence, and the interspersal of inflectional and derivational material. Aymara (Chapter 22), eponymous member of the Aymaran family of Bolivia and Peru, represents a highly agglutinative, exclusively suffixing system. Nevertheless, it also displays a convincing instance of subtractive morphology, in that the accusative case is formed by apparent vowel deletion. Nen (Chapter 23) of the Morehead-Maro Family of Southern New Guinea, has a particularly elaborate system of distributed exponence in which verbal argument features are constructed through information spread across prefixes, suffixes, stems, and pronouns. Its aspectual system is based on the typologically unusual distinction of ingressive versus non-ingressive. Inflectional exponence in Shilluk (Chapter 24), a language of the West Nilotic branch of Nilo-Saharan (spoken in South Sudan), is characterized by stem-internal alternations of (p. 7) length, tone, and ATR (advanced tongue route) values, along with stem-final consonant alternations. Since its affixal morphology shows a great deal of homophony or polyfunctionality, stem-internal marking plays an increasingly dominant role in the system.


First and foremost I would like to express my appreciation to the contributors to this volume, who have done most of the work. The collaboration has been a rewarding one. I also thank Penny Everson for her invaluable assistance in the preparation of the manuscript. Finally, none of this would have been possible without the support of the European Research Council (grant ERC-2008-AdG-230268 MORPHOLOGY), which is gratefully acknowledged. (p. 8)


(1) While morphology obviously embraces more than just inflection, in practice only inflection is ever adduced as a reason for postulating an autonomous morphological component.

(2) Word class in this sense is a morphological notion that may well cut across syntactic categories; for example, the verbal paradigm in Latin is traditionally construed as containing nominal forms, namely participles and gerunds.