Introduction

In the mid-sixties, the philosopher G.H. von Wright proposed to analyse change in terms of a transition operator T (‘and next’) from a certain state described by the proposition ¬φ to a certain state described by the proposition φ. Independently, the linguist J. Gruber proposed to analyse change in terms of a ‘moving’ entity, the Theme, which ‘goes’ from a Source- to a Goal-position.

In both approaches, change is analysed in terms of an initial state and an end state which are related to points of time with respect to which these states are said to hold. However, there is a fundamental difference between the two. On the Von Wright approach, the interval being defined in terms of the initial state and the end state leaves the stretch between them as a sort of no man's land. No information is given by T, so a natural question is: what happened between the two states? All sorts of detailed questions arise, such as: at what point does the transition take place if there is such a point? What does the interval structure look like? What about subintervals? Is there a smallest interval?, etc. (Kamp 1979, 1980; Van Benthem 1983a). This approach put the concept of homogeneity at the centre of attention in the seventies and early eighties.

We have seen in chapter 2 how Bloomfield's concept of the morpheme was transformed by his immediate successors (above, §2.3). But we have yet to consider contemporary developments in syntax. Once again a part of Bloomfield's theory was taken from its context and made the foundation of the whole. In morphology it was the morpheme, so that an account of morphology, insofar as it remained distinct from syntax, was reduced to the identification of morphemes and their alternants. In syntax it was, above all, the principle of immediate constituents. From the late 1940s syntax had basically two tasks, one to establish the hierarchical structure of sentences and the other to sort the units of this hierarchy into classes with equivalent distributions. For the same period also saw the firm adoption of distributional criteria. Not merely did the study of language start from form rather than from meaning; but the investigation of form was separated strictly from that of meaning, and necessarily preceded it. It was in syntax that this programme was particularly attractive, and met with the fewest doubts and criticisms.

Of the sections which follow, §3.1 deals with the origins of distributionalism, up to and including the classic Post-Bloomfieldian treatments in the 1950s. In §3.2, we will see how these ideas gave rise, again in the 1950s, to the concept of a generative grammar. This section will concentrate, in particular, on Harris's Methods (1951a) and Chomsky's first book (1957). We will then turn to the development of the constituency model, which culminated in the formalisation of phrase structure grammar (§3.3).

The concept of the morpheme is central to much of twentieth-century linguistics, and its history, especially in America, has been told before. But there are reasons for looking at it again. One is that, in interpreting the account in Bloomfield's Language, few commentators have made detailed reference to the earlier book, An Introduction to the Study of Language, which it replaced. But I will try to show that the comparison is interesting, for morphology and for grammatical theory generally. Another reason is that many have accepted far too readily the critique of Bloomfield's theory by American scholars since the 1940s. I will argue in §2.3 that it was in part mistaken, and that the weaknesses of his treatment, though real and serious, lay elsewhere. A third reason is that, although morphology was largely neglected in the heyday of generative phonology, a new chapter in its history began in the 1970s. I will therefore try to explore the connection between the work of the past twenty years and what had gone before. Finally, I would like to place the theory of the morpheme in what I believe to be its historical context. This can be seen as part of a wider tension or conflict, discernible sporadically throughout the century, between the concepts and techniques of grammar inherited from earlier Western traditions, and the doctrines or (if we prefer) the insights of structuralism.

In morphology, the tension lies particularly between two views of the status, or the primary locus, of grammatical categories.

Many readers will be familiar with the classic historiographic study by Walter Carruthers Sellar (Aegrot. Oxon.) and Robert Julian Yeatman (Failed M.A. etc. Oxon.), in which they set out ‘all the parts you can remember’ of the History of England (Sellar & Yeatman, 1930). What ‘Every student can remember’ of the history of linguistics is not perhaps so bad, and sadly less hilarious. But it would not be difficult to put together an account of ‘1957 and All That’, in which developments in the twentieth century are quite seriously garbled.

It would contain at most two ‘memorable dates’. One is that of the publication, in 1957, of Chomsky's Syntactic Structures, in which Structuralism, or (according to some authorities) American Descriptivism, was overthrown. The other date, which careful research might well reveal not to be memorable, is that of Saussure's Cours de linguistique générate. Before this, at the beginning of the century, linguists were only interested in the history of languages. But according to Saussure, who is known as the Father of Modern Linguistics, the subject had to be synchronic, and we had to study ‘la langue’, which is just an arbitrary inventory of signs. This was at first a Good Thing, since it led to a lot of important work especially on American Indian languages. But in the long run structuralism was a Bad Thing. One reason is that the structuralists did so much work with American Indians that they came to believe that languages could differ from each other in any way whatever.

Parts of this book are based on earlier publications. Chapter 2 substantially reproduces a paper with the same title that is about to appear in the Transactions of the Philological Society (Matthews, 1992); I am grateful to the Secretary for Publications for allowing that earlier version to go ahead. In the first two sections of Chapter 3 I have incorporated some material from my contribution to a Festschrift for R. H. Robins, edited by F. R. Palmer and Th. Bynon and published by Cambridge University Press (Matthews, 1986). A preliminary and much shorter version of Chapter 4 appeared with a different title in An Encyclopaedia of Language, edited by N. E. Collinge and published by Routledge (Matthews, 1990a). It has been rewritten almost entirely, but I am grateful for their permission to incorporate material from it.

John Lyons has very kindly read and commented on the final typescript. I dedicate the book to my wife, Lucienne Schleich; she has not only commented on most of it but, more importantly, she has taught me the self-discipline needed to write it.

Chomsky's first article appeared in 1953; forty years later, when this book is published, he will barely have reached a normal age of retirement. It would be astonishing if, over these four decades, his views had not changed, even on general topics and even touching matters central to his thought. In the beginning, for example, he believed that the study of meaning was separate from that of grammar, and concerned with the ‘use’ of sentences. Some years later, he thought it obvious that sentences had ‘intrinsic meanings’ that were determined by grammatical rules. At still later stages, that belief was gradually reversed. At no moment, however, have his views changed comprehensively and suddenly. They have changed as the clouds change in the sky, or as the key changes in a classical sonata movement. Nor, in some ways, have they changed entirely. As in a sonata movement, new ideas have often been at once new and a development of perhaps a fragment of what has preceded.

The aim of this chapter is to try and trace the evolution of Chomsky's general ideas. It will be well to acknowledge at the outset that this is not easy, and that any account that might be given, barring the most slavish and unpenetrating chronicle, is liable to seem in part misleading to others who have read his works differently. One reason is that Chomsky's career is not over: his ideas are still developing, and, even if scholars are less eager than before to publish their reactions, anyone with a serious interest in the nature of language must still try to decide for or against them.

Introduction

This chapter poses the question: ‘What is the head of modified nominal structures such as good students, these students or many students}’ The answer I shall give here is one inspired by classical mythology: namely, that like the ancient Roman god Janus, such structures are double-headed. The theoretical framework used here will be that of Government-Binding theory; the specific descriptive claim being made is that modified nominals incorporate multiple phrasal projections (with each modifier heading a separate projection), and that each phrasal ‘layer’ of the structure comprises both an immediate head (the modifier) and an ultimate head (the modified N): thus, for example, the immediate head of the expression ‘good students’ is the Adjective good, but its ultimate head is the Noun students. The analysis will proceed in a bottom-up fashion: accordingly, I start by looking at the innermost NP ‘core’ of nominals.

The internal structure of Noun Phrases

Consider the internal structure of nominals such as the following:

1. (1) a. ministry of defenceinstructionsto all employees

2. b. governmentcriticismof the press

3. c. Labour Partypolicyon defence

4. d. military policeinvolvementin torture of prisoners

5. e. university managementallegationsof a concerted student campaign of disruption

6. f. European Communitydemandsfor monetary union

7. g. studentassessmentof lectures

8. h. Department of the Environmentplansfor a new motorway

The overall string in such examples is traditionally considered to have the categorial status of a Noun Phrase: its head in each case is clearly the capitalized Noun, since the number properties of the capitalized Noun determine the number properties of the overall nominal, so that, for example, a nominal such as Ministry of Defence instructions…

Introduction

In this chapter I wish to consider the implications of juxtaposing three strands of recent syntactic research that have tended to remain independent of each other. Although there are undeniably differences of emphasis and occasionally of principle, it is part of the contention of this chapter (a) that there is much to be gained from an attempt to bring them together, and (b) that there is less of real substance keeping them apart than is perhaps sometimes thought. The approaches in question are:

1. The line of typological research inaugurated by Nichols (1986), in particular her fundamental distinction between head-marking and dependent-marking languages.

2. The focus in the recent generative literature on the properties of functional categories and their projections (Chomsky, 1986a; Abney, 1987; Speas, 1990 and a whole host of other references).

3. The study of the processes of grammaticalization as a mechanism of syntactic change (Heine, Claudi and Hunnemeyer, 1991; Heine and Traugott, 1991; Hopper and Traugott, forthcoming).

I will begin with a brief characterization of each.

Head-marking versus dependent-marking

This typology assumes a theory-independent and cross-linguistic agreement as to which is the head and which the dependent in any given syntactic construction, and then classifies languages according to whether the head-dependent relation is marked on the head or the dependent. Thus, compare the following:

1. Maltese bin Alia

2. son of God

3. Latin filius Dei

4. son of God

The Maltese noun iben ‘son’ has a special form bin, the so-called construct state, which is required when it has a nominal dependent. The dependent nominal Alia ‘God’ occurs in the same form as it would if it were an independent element.

Introduction

Two key issues in the discussion of heads are: how is ‘head’ to be defined, and are there clear linguistic generalizations that provide evidence for this notion? The debate between Hudson and Zwicky in the Journal of Linguistics (Hudson, 1987; Zwicky, 1985) illustrates the difficulties inherent in answering the first question. There is disagreement over which grammatical properties define, or correlate with, heads, and hence over the precise set of categories that are supposedly instances of this generalization. With regard to the second question, one grammatical area that has been cited as relevant is word order (see, for example, Zwicky, 1985: 10–11). It is often proposed that heads are consistently positioned either before or after non-heads within their respective phrasal categories. Vennemann (1974) formulates a Natural Serialization Principle for this regularity, and reformulates it in Vennemann (1984) in terms of the typological distinction between ‘pre-specifying’ (that is, modifier before head) and ‘post-specifying’ (head before modifier) languages.

This chapter will consider what word-order universals as currently formulated suggest about the notion ‘head of phrase’. I will argue, with Hudson, that there is a significant generalization here, but that it is not ultimately the kind of generalization he proposes. Indeed, the notion ‘head’ can probably be dispensed with altogether, as a property of Universal Grammar. Instead, I believe the underlying generalization involves a principle of parsing: the relevant categories provide unique ‘Mother Node Construction’. They enable the language user to recognize the existence and nature of the syntactic grouping of words to which the word and category in question belong, and so inform the listener about abstract syntactic structure on-line.

Introduction

A major focus of the debate on headedness has been the problem of determining the head in different constructions and of establishing acceptable criteria to enable us to do so. The data have been taken mainly from English, and so this account, by contrast, extends the investigation to a language with a much richer morphological system than that of English, namely Russian. We shall concentrate on numeral expressions in Russian, where the head-dependent relation has long been known to be problematic (see, for example, Isačenko, 1962: 529). We shall examine them in the light of the criteria for heads proposed by Zwicky (1985) and by Hudson (1987). At first sight it seems that no single head can be identified for these constructions; rather, the properties of the head appear to be shared between different elements, which would fit with Zwicky's approach. However, given current assumptions about lexical entries and feature distribution, these constructions can be analysed as being rather less exotic than they first appear, and as having a consistent head, as Hudson would predict. While attempting to remain as theory-neutral as possible, we shall develop the analysis to see whether the idea of a single element having all the head properties can be maintained. It is in focusing on the question of headedness that this chapter differs from most previous accounts of Russian numeral phrases. We shall see that there are two consequences. The first is that we still need to recognize that headedness is a gradient notion: a particular element may have head-like characteristics to a greater or lesser degree, and that these may vary according to external factors (notably, case assignment).