A Serial verb construction (or SVC) is one in which a single clause contains two or more verbs, neither of which is an auxiliary. The term “serial verb” has been used by different authors in slightly different ways, and linguists sometimes disagree about whether a particular construction in a given language is “really” a serial verb or not. We will not attempt here to formulate a precise definition that will clear up all these gray areas. Rather, we will focus on describing the most important features of the “prototypical” SVC, i.e., characteristics of those serial verb constructions which everyone agrees belong to this category. As a way of introducing this discussion, let us consider a type of causative construction which is similar in many ways to those discussed in chapter 8, but also different in ways which make it relevant to the current chapter.

Causative constructions in French

Causative constructions in Romance languages are periphrastic, in the sense that they contain a verb meaning ‘cause’ which is morphologically independent of the base verb. However, they can also have many of the properties associated with morphological causatives. In particular, some of these constructions can be shown to be monoclausal, even though they contain two verbs. In this sense they are similar to serial verb constructions.

French, like English, lacks a morphological causative; but it has two different types of periphrastic causative.

This book provides a framework for analyzing and describing grammatical structure, building on what linguists have learned about language in general while paying careful attention to the unique features of each particular language. Its primary focus is on syntax (sentence structure), but it also deals with aspects of meaning, function, and word structure that are directly relevant to syntax.

This is a book about syntactic analysis, rather than syntactic theory. I have adopted a simplified version of Lexical Functional Grammar (LFG) as the analytical framework for the book, but I have tried in each chapter to emphasize linguistic phenomena over formal notation. The analyses presented here are very much in the spirit of LFG, but the notation employed is modified and simplified compared to that of standard LFG. Those readers who want a more complete introduction to LFG as a formal system are encouraged to consult Bresnan (2001), Dalrymple (2001), or Falk (2001).

This book is written at a level which should be appropriate for advanced undergraduate or beginning graduate students. It presupposes some familiarity with basic linguistic concepts and terminology, but no previous background in formal syntax. The contents can be covered fairly easily in a typical semester-length course. The book does not assume that its readers have native-speaker intuitions about English. So, for example, in some places the meanings of English idioms are explained, alternative possible interpretations of certain constructions are explicitly spelled out, etc.

A Relative clause construction is a noun phrase which contains a clausal modifier. For example, the subject NP in sentence (1) consists of a determiner (the), the head noun (woman), and a modifying clause (I love). The modifying clause is introduced by a Relativizer (that).

1. (1) [The woman [that I love]s′]NP is moving to Argentina.

Note that the modifying clause in this example seems incomplete: it lacks a direct object, even though its verb (love) requires one. Nevertheless, this sentence is perfectly grammatical. We can give an intuitive explanation for this fact by suggesting that the head noun woman is “understood” to be the object of love. Another way of expressing this intuition is to say that the modifying clause contains a “gap,” while the head noun is interpreted as the thing which fills this gap, making the sentence complete.

In English, this kind of “filler–gap” structure is also found in content (or Wh-) questions and several other constructions, such as clefting and topicalization (i.e., the Contrastive Topic construction). Constructions of this type have some unique and interesting properties, which have stimulated a great deal of syntactic research. One characteristic property of these constructions is that the “filler” may be very far from the “gap.” In English Wh- questions, for example, the question word may occur an arbitrarily long distance away from the verb which assigns its semantic role and grammatical relation.

In chapter 10 we saw that case and agreement markers are not always in a one-to-one correspondence with grammatical relations. For example, we discussed a number of languages in which certain verbs require special case marking for their subjects or objects. In other languages, the marking of subjects depends on the transitivity of the clause: transitive subjects are marked one way, while intransitive subjects are marked another way. When, as is frequently true for such languages, intransitive subjects are marked in the same way as transitive objects, we say that the language is Ergative.

Ergativity in case or agreement systems is not uncommon among the world's languages. More rarely, however, we encounter languages in which the kind of subjecthood tests discussed in chapter 10 reveal an “ergative” pattern in the syntax. This “syntactic ergativity” will be the main focus of this chapter. We will suggest that syntactically ergative languages differ from other, more familiar, languages in their pattern of linking grammatical relations to semantic roles. But first, let us review the properties of a simple ergative case system.

Morphological ergativity

As we saw in the preceding chapter, the subject of a German clause normally takes nominative case. This is true whether the clause is transitive, as in (1a), or intransitive, as in (1b). Direct objects take a different case marker, namely the accusative (1c).

Probably no one has ever before said or heard the following sentence, yet any normal adult speaker of English will understand it:

1. John Adams could have been elected to a fourth term as President, if his step-sister had not been so ugly.

In the same way, a speaker of any language will say and hear many sentences during the course of a normal day which he has never said or heard before. Moreover, other speakers of the same language will not only recognize these original creations as being well-formed sentences but will also (usually) understand what they mean.

These observations tell us something important about the nature of language. A person who knows how to speak a language does not have to memorize every possible sentence in that language. Rather, speakers produce sentences Creatively. Some common phrases and sentences may be repeated so often that they are memorized as a single unit, e.g., idioms and proverbs. But, for the most part, we do not memorize sentences; rather, we construct them when we need them, to express a particular idea.

This creative use of language is possible because the patterns of a language are determined by a set of Rules. A speaker who (unconsciously) “knows” these rules can use them to create and understand any number of new sentences.

In this chapter we will examine some examples of “irregular” or atypical case marking. Case is a morphological feature used to indicate the function of an NP within its clause. But, as we will see, case marking is not always a reliable guide for determining the grammatical relation of a particular NP: an NP may be a grammatical subject or object even though its case marking suggests something quite different.

In fact, it is not uncommon to find this kind of “mismatch” (or irregular correlation) between morphological features and syntactic functions. For this reason, it is important that grammatical relations be identified on the basis of syntactic evidence, and not on purely morphological grounds. In this chapter we will be primarily interested in the mismatch between case marking and grammatical relations. Examples of this kind will help us to clarify the kinds of syntactic evidence which can be used to identify grammatical relations (subjects in particular) when there is reason for uncertainty.

We begin by looking at examples from German and Icelandic which show how the regular case-marking rules for subjects and objects can be overridden by the lexical requirements of a particular verb. As a way of introducing this issue, we will first consider the marking of prepositional objects in German. In the second section of this chapter we will consider the “dative subject” construction, taking examples primarily from South Asian languages.

As we saw in chapter 1, phrase structure diagrams are used to represent several different kinds of information about the structure of a sentence: (i) word order; (ii) constituent boundaries; and (iii) the category of each word and constituent in the sentence. It is important to remember that these tree diagrams are just pictures which are supposed to represent certain linguistic properties of real sentences. Before we can draw the pictures, we first have to understand the linguistic reality which they are intended to represent.

For example, we stated that a Constituent within a tree structure corresponds to “all and only the material which is dominated by a single node.” But this is primarily a statement about how we draw trees. In order to apply this definition, we first need to be able to find constituents in real language data, i.e., to determine which groups of words function as a “unit” in a particular sentence. Our decisions about how to draw the tree are based on observable facts about the language.

This is a specific instance of a general point which seems obvious, but is nevertheless worth emphasizing: our analysis of the grammatical structure of a language must be based on linguistic evidence. In this chapter we will discuss various kinds of evidence which can help us answer basic analytical questions about constituent structure.

Consider the following three sentences:

1. (1) a My cat died.

2. b The Mayor caused my cat to die.

3. c The Mayor killed my cat.

Sentence (1a) is a simple clause which describes an event involving one argument, namely my cat. Sentence (1b) is a biclausal construction involving two participants, and describing two distinct events. The complement clause in (1b) describes the same event that was expressed in (1a), while the matrix clause refers to a different event in which the Mayor performed some unspecified action. Clearly there is a logical connection between the two events: one is the result of the other. In other words, the first event (the Mayor's action) is seen as the cause of the second (my cat dying).

Sentences (1b) and (1c) are similar in meaning, but not perfectly synonymous. Sentence (1c) seems to imply that the Mayor personally killed my cat, while (1b) would be more appropriate if the cat's death was an indirect result of some action on the Mayor's part (e.g., ordering that all the dogs in the dog pound should be released). Nevertheless, there is a substantial overlap in meaning between the two sentences. Both imply a cause-and-effect relationship between two events; the causing event in both cases is something the Mayor did; and the effect (or result) in both cases is the event described in (1a).