Search

Syntax is the system of rules that we subconsciously follow when we build sentences. Whereas the grammar of English (or other languages) might look like a rather chaotic set of arbitrary patterns, linguistic science has revealed that these patterns can actually be understood as the result of a small number of grammatical principles. This lively introductory textbook is designed for undergraduate students in linguistics, English and modern languages with relatively little background in the subject, offering the necessary tools for the analysis of phrases and sentences while at the same time introducing state-of-the-art syntactic theory in an accessible and engaging way. Guiding students through a variety of intriguing puzzles, striking facts and novel ideas, Introducing Syntax presents contemporary insights into syntactic theory in one clear and coherent narrative, avoiding unnecessary detail and enabling readers to understand the rationale behind technicalities. Aids to learning include highlighted key terms, suggestions for further reading and numerous exercises, placing syntax in a broader grammatical perspective.

Summary

We humans are surrounded by technology. We have machines for almost everything and computers allow us to achieve things that were long deemed impossible. We live in a day and age where dreams can become reality overnight due to technical innovation, and the amount of information that we have access to via a tiny machine in our pockets is simply astounding. There is talk right now about flying people to Mars and growing tomatoes there. Yes, we humans are a smart bunch.

Despite all this, there are some things we are still unable to do, and some machines we simply cannot build. And some of these failures have to do with language. A machine that can translate one language into another language perfectly? No, we don't have it (and please don't insult us by referring to Google Translate). Okay, how about something more modest, like a machine that can for any combinations of words in a single language (say, English) say whether it is a good sentence or not? It is perhaps hard to believe but even that is still out of our reach. Language, as it turns out, is an evasive and slippery creature.

At the same time, it is clear that such a machine, capable of stating for every English sentence whether it is grammatical or not, does exist. In fact, we have about 360 million of those machines on our planet. They are called native speakers of English. These speakers have at their disposal the knowledge of their mother tongue, English, and this knowledge can generate zillions of distinct combinations of English words and evaluate each of them, whether old or new, as being either a good sentence or not. Probably you have never heard someone say Syntax is one of the most fascinating topics in linguistic theory, but if you are a native speaker of English you know immediately that the sentence is grammatically correct (and hopefully after reading this book you will also find it to be correct content-wise). So these native speaker brains can do something that we cannot imitate with any man-made machine. The fact that we cannot mimic such everyday human language behaviour shows us that there is something worthwhile studying.

Summary

CHAPTER OUTLINE

We have seen that words or constituents can remerge. Remerge, also known as movement, explains a lot of things. Thanks to Remerge we are able to understand that a constituent can appear in a position far removed from the case and θ-role assigner on which it depends. And, by assuming that subjects start out in the VP and are subsequently remerged, we are able to treat all syntactic dependencies on a par, including the previously problematic nominative case relation between a subject and Fin. But what we don't yet understand is why elements must remerge. Why do subjects remerge into FinP, and Wh-phrases into CP? In this chapter we address this question, and we will conclude that the same feature-checking mechanism that we have already used to characterise syntactic dependencies provides us with the trigger for such instances of Remerge.

Insight: Agreement Triggers Remerge

In the previous chapter, we established that constituents – heads and phrases alike – can be remerged into the structure we build, thereby giving rise to the effect that we know as movement. By adopting this notion of Remerge we were able to understand a number of phenomena that would otherwise remain rather mysterious. In addition, it allowed us to maintain generalisations that would otherwise be lost, such as our unified characterisation of syntactic dependencies in terms of [F]–[uF] feature checking.

The fact that we can analyse certain phenomena as involving Remerge, however, does nothing to explain why these instances of Remerge actually take place. Take two significant Remerge operations that we have considered: (i) Remerge of a Wh-constituent to a clause-initial position and (ii) Remerge of the subject from VP to FinP.

(1) a. Which yellow chair has Adrian always liked <which yellow chair>?

b. The teachers are all <the teachers> dancing on the table.

Assuming that Remerge has taken place in (1a) allowed us to maintain our restrictions on case agreement, and to assume that θ-role assignment is strictly local, and always takes place within the VP. These reasons also accounted for remerger of the subject in (1a) from spec-VP to spec-FinP, and for the presence of the floating quantifier all between the auxiliary and main verb.

Summary

CHAPTER OUTLINE

In the previous chapter we encountered a particular syntactic dependency: case. Even though sentences containing constituents with the wrong case marking would be semantically good, syntax does not allow them. In this chapter we will see that case is not the only kind of syntactic dependency that can be identified. There are more constructions that, from a semantic perspective, would be good, but that syntax does not like. For instance, sentences like *I walks (instead of I walk) or *She walk (instead of She walks) are bad. Other examples are sentences like *John hopes that Mary loves himself (instead of John hopes that Mary loves him), or *I love me (instead of I love myself). These sentences are clearly ungrammatical, even though it is not that hard to figure out what they could mean. The big question that arises is whether all these syntactic dependencies are different in nature, or underlyingly the same. The latter would be the best outcome, since in that event there is only one mechanism in syntax that we need to understand rather than several. Indeed it turns out that all these syntactic dependencies are the result of a single mechanism: agreement.

Insight: Agreement Reflects Syntactic Dependencies

The previous two chapters introduced two main constraints on what you can create with the Merge operation: θ-theory and Case theory. θ-theory is essentially a semantic constraint; verbal categories that merge with too few or too many arguments yield degraded sentences, since the meaning of the verb dictates how many arguments it should merge with. θ-theory has an effect on syntax, since it partly determines the structural size, but it is not a syntactic constraint. Case theory, by contrast, is a syntactic constraint. Sentences that are perfect from a semantic point of view (She loves I instead of She loves me) are ruled out by the syntax. Every DP should be assigned case, and particular structural positions are responsible for particular case assignments.

A question that may now arise is whether there are more syntactic dependencies than just those involving case assignment. The answer is yes.

Summary

In this book, we have moved from looking at the grammar of English as a set of basically arbitrary rules toward looking at the grammar of English as a set of rules that are the result of a restricted set of principles underlying them. To put it differently, we have uncovered several fundamental patterns underneath the surface rules.

Here is an example. The fact that in English a Wh-constituent must be in clause-initial position is a surface rule that is caused by the underlying fundamental principle for interpretable features to c-command their uninterpretable counterparts. The former (surface) rule is not fundamental for English, the latter is: the former is a consequence of the latter and not the other way around. And the latter principle captures not just the surface rule that Wh-constituents appear in clause-initial position, but it captures all syntactic dependencies.

Another one. The rule that in English you must say eat a sausage and not a sausage eat is not an arbitrary fact about verbs and objects in English but follows from the more fundamental choice in English to linearise complements after their heads. This is not just true for verbs and objects but also for determiners and nominal phrases, and for complementisers and clauses. It is, in fact, the rule within English phrases in general.

Many more examples can be given. Syntacticians try to capture general, underlying patterns on the basis of the surface patterns we observe, and formalise these into a theory. The theory we developed in this way for English meets a couple of requirements we can set for any scientific theory:

Summary

CHAPTER OUTLINE

So, let's start. In the introduction we stated, very ambitiously, that by studying the rules underlying proper English sentences we may ultimately be able to better understand what is going on in the human mind or brain. So how do we study those rules? We will begin by looking at simple sentences and reveal an important insight: syntactic rules do not apply to particular words, but rather to particular categories that words belong to, e.g. nouns and verbs. That simple but fundamental insight already reduces the number of potential syntactic rules in a drastic way, and therefore saves us a lot of work. Moreover, we will see that words belong to a particular category because they carry features that are purely syntactic, and that must be distinguished from features that tell you how to pronounce those words or that tell you what they mean. These features will play a very important role throughout the book, and, for starters, they will help us make a first syntactic discovery: there are words that are never uttered.

Insight: Words and Categories

Syntax tries to understand why certain sentences (like Mary loves Suzanne) are good English sentences, whereas other ones (such as loves Mary Suzanne or Mary snores Suzanne) are bad English sentences. What causes them to be either good or bad?

Language consists of words. Words are elements that can be combined into sentences. In a way, you can think of words as the building blocks (like Lego blocks) with which sentences can be built. A simple sentence like Mary loves Suzanne consists of three words. The first idea therefore might be to say that there are certain rules in language (in this case, the English language) that allow the words loves, Suzanne and Mary to be combined in a particular order.

In essence, then, you need rules. If there weren't any rules, nobody could explain the correctness, often called the grammaticality, of some sentences and the ungrammaticality of others. There would be no way of stating the distinction if there were no rules. But how specific must these rules be?

Summary

CHAPTER OUTLINE

We saw in the previous chapter that every syntactic dependency reflects a relation between an interpretable and uninterpretable feature of the same type. In this way, we were able to generalise over case assignment, agreement relations and binding phenomena. However, we were not at that stage able to provide a uniform analysis for [F]–[uF] agreement, since nominative case agreement did not really fit in: the [uFin] feature on the subject in spec-FinP c-commands the [Fin] feature on the Fin head (and not the other way around), whereas other dependencies show the opposite c-command relationship. We have seen that an attempt to simplify the system usually leads to new insights. We will therefore hypothesise that we have not yet been able to account for agreement in a uniform way because we are still missing something. This chapter introduces this missing piece of the puzzle, which is a topic in its own right. It involves the recognition that constituents sometimes appear in positions in which we do not interpret them. They seem to move around. Syntactic movement, we will show, is not a new property of syntax. Rather, it follows immediately from the building procedure we already have: Merge. The existence of movement phenomena is therefore a prediction, and a confirmed one, made by the Merge hypothesis.

Insight: Constituents Can Move

So far, we have created structures by binary application of Merge. What does this mean? It means that you take two constituents and put them together. And if you want to build something bigger, you take a new constituent and add that to the structure you previously created. So if you have a sentence like Adrian has always liked this yellow chair, you take yellow and chair and merge them, and you merge the result yellow chair with this. Then you merge the DP this yellow chair with liked; this result you merge with always, you then take has and merge it with what you have. Finally, you merge that product with Adrian and you get Adrian has always liked this yellow chair.

Summary

CHAPTER OUTLINE

In chapter 2, we saw that the Merge operation, elegant as it is, can still create both grammatical and ungrammatical structures. We therefore concluded that we need constraints that filter out these ungrammatical structures. θ-theory was the first serious constraint we introduced. The nature of this constraint is semantic: we saw the number of arguments present in the syntax must match the semantics of the verb. Sentences like Edith assigns or John loves Mary Paul (where Mary and Paul are two different people) are now correctly filtered out. It is quite easy to see, however, that more work needs to be done. Let us give you a very simple example. The sentence Him calls she is ungrammatical, despite the fact that the verb can assign both its θ-roles to the arguments that are present (i.e. the AGENT role to him and the PATIENT role to she). In terms of the meaning, then, nothing is wrong, but something is not quite right with the form of the sentence. This chapter will explore in detail how to make this explicit, and how syntax will ensure that the structures it creates have the right form. This second filter on the Merge operation is called Case theory. We will see that the Case theory presented in this chapter has far-reaching consequences for the syntax of English sentences.

Insight: Case as a Filter on Syntactic Structures

Although θ-theory is a necessary filter on constructions created by the Merge operation, at least three problems emerge if we say that θ-theory forms the only constraint on Merge.

The first problem concerns θ-mismatches, examples in which the number of arguments does not match up with the number of θ-roles that the verb needs to assign. It turns out that an unassigned θ-role is not as bad as an argument without a θ-role. For instance, transitive verbs that are used intransitively can be quite bad (as expected by θ-theory), but sometimes they can survive. Edith assigns could be said in a situation in which the manager, on asking who assigns tasks to new employees, is told that it is Edith who does the assignments. The same holds for the verb to kill.

Summary

CHAPTER OUTLINE

In the previous chapter, we showed how morphology provides concrete, pronounceable forms for the abstract feature bundles that syntax merges. Once the grammar knows which morpho-phonological forms to use, it starts to care about how exactly to pronounce them. In this chapter, we will not concern ourselves with the actual way these morpho-phonological words are realised acoustically. The reason is that there is simply no connection with syntax here. The concept CAT in English is pronounced as K-A-T (or in the International Phonetic Alphabet as /khæt/), the pronoun he is pronounced as H-E or /hi:/. This is in no way related to the syntactic structure, and therefore of no interest to a syntactician. There is one aspect of sound, however, that a syntactician should care about, and which therefore deserves a closer look. It concerns the fact that the morpho-phonological forms that morphology has provided (i.e. the words of the sentence) have to be pronounced in a particular order (call it the word order). This order is not random but obeys certain rules. This raises the question of how to formulate those rules. It turns out that we run into a bit of a paradox here. On the one hand, it seems clear that the word orders we end up with are not entirely independent of the structures created by syntax. On the other hand, these word orders are not fully determined by the syntactic hierarchy either. After all, a hierarchy is not a linear order. This means that, in order to understand how linearisation works, we have to study the relation between syntax and phonology carefully. Which part of word order is due to syntax, and which is due to phonology? This is the topic of this chapter.

Insight: Syntax is not about Word Order

Syntax is not about word order? Yes, this is what we mean. This statement may come as a bit of a surprise because many dictionaries will define syntax in exactly that way: syntax is the study of word order. Are these dictionaries all wrong, then?

Summary

This book is based on the syntactic approach known as minimalism and follows the generative paradigm as devised by Noam Chomsky. The literature on minimalist syntax is enormous and it is impossible to present a complete overview of all the relevant literature. In the main text we have spelled out what the major sources are for each chapter, as well as options for further reading. Please realise, though, that often we have simplified particular analyses for didactic or presentational purposes or represented them from another angle. Also realise that many of the references may not be easy to read (Chomsky's writings in particular can sometimes be very hard). For more accessible texts, we refer you to certain syntactic handbooks (e.g. Den Dikken 2013, Everaert et al. 2013 and Alexiadou & Kiss 2015) or the introductory works of our fellow syntax textbook writers, most notably Adger (2003), Haegeman (2006), Radford (2009), Tallerman (2011) and Carnie (2013).

Summary

CHAPTER OUTLINE

We have seen that words in a language are not an arbitrary set of items, but that every word belongs to a particular syntactic category, such as ‘noun’, ‘verb’, or ‘complementiser’. In this chapter, we are going to look at combinations of categories, for instance the combination of an adjective and a noun. We can make a very simple observation here about such a combination: an adjective and a noun together behave just like a noun without an adjective. In other words, the noun is more important than the adjective. Although this observation is simple, the consequences will turn out to be enormous: the most important is that phrases and sentences should be analysed as structures or, to be even more precise, hierarchies. This insight is perhaps the most fundamental in syntactic theory. This chapter will develop this idea and present the most important predictions that it makes. These predictions turn out to be correct and therefore provide strong evidence in favour of this idea.

Insight: Constituents Are Headed

As established before, nouns can be preceded by one or more adjectives. We can, for instance, say sausages and delicious sausages. In the latter case we have combined one unit, delicious, with another unit, sausages, thereby creating a third, bigger, unit: delicious sausages. The word linguists use for such units is constituent, so we will use this term from now on.

What can we say about the behaviour of the constituent containing a noun and an adjective? Would it behave like a noun, like an adjective, or like something different? In order to investigate this, let's take the constituent delicious sausages and see what grammatical properties it has. In other words, let's see in what kind of syntactic environments this constituent of two words can stand.

First, delicious sausages can be combined with another adjective, like expensive, giving us expensive delicious sausages. Second, we can put the article the in front of it: the delicious sausages. Third, we can change delicious sausages (which is plural) into a singular expression: delicious sausage. Fourth, we can have it followed by some prepositional expression, such as from Italy, as in delicious sausages from Italy.

Summary

Introducing Syntax

Whereas the grammar of English (or other languages) might look like a rather chaotic set of different patterns, syntactic theory has revealed that these patterns can actually be understood as the result of a small number of grammatical operations. Unravelling these is the science of syntax. This textbook describes state-of-the-art syntactic theory by addressing how and why certain combinations of words are ‘proper’ English sentences whereas other combinations are not. What is the mechanism behind that? What grammatical rules does English have and why? How is grammar related to meaning and to how sentences are expressed?

In this book we guide students through a variety of intriguing puzzles, striking facts and novel ideas, and let them discover the beauty of syntactic theory in both a bottom-up (data-driven) and a top-down (theory-driven) fashion. This book is primarily intended for students for whom this is a first (and hopefully not last) encounter with syntax and/or linguistic theory. We will primarily focus on the important insights that have been achieved in contemporary syntax. Introducing Syntax will offer students all the necessary tools to do this, without going into unnecessary technical detail.

Introducing Syntax is not the only available textbook on this topic. Why, then, the need for a new one? Introductory courses to (English) syntactic theory generally face three key challenges:

Summary

CHAPTER OUTLINE

In the previous chapters, we focused on the relation between syntax and the way syntactic structures are expressed: how do concrete morpho-phonological forms get inserted into the syntactic representation, and how do these forms get linearised in the order we hear them in? But sentences are not only expressed; they also have a meaning. Semantics is the grammatical component that assigns an interpretation (i.e. a meaning) to a sentence. How does it do that? Crucially, like morphology and phonology, semantics is also based on the output of the syntax: the way words and morphemes with their interpretable features are merged determines the meaning of a sentence. This sounds like a pretty straightforward procedure, but we will see that the relation between syntax and semantics is not always transparent, as was also the case for syntax and morphology. In order to get to the right interpretations of sentences we sometimes need more syntax than syntax seems to give us.

Insight: Compositionality

How does a sentence get the meaning it has? This is a question that has interested philosophers and linguists for centuries. And even though there are many complex issues at stake, at least on two points consensus has arisen. These two points may sound obvious at first sight, but appear to be very powerful in explaining all kinds of semantic effects.

The first point is that the meaning of a sentence is determined by the meaning of the words or morphemes it consists of. If you have the sentence John love-d Mary, the meaning of the entire sentence follows from the meaning of components like John, love, -d and Mary. This sentence cannot all of a sudden mean that some guy hated Bill. No meaningful elements are removed or added when we interpret this sentence. In that sense, sentence meaning is like a linguistic fruit salad. If you mix bananas, strawberries and apples, the salad is not going to taste like oranges. Now, you may wonder, though, what exactly we mean when we say that semantics is interested in the meaning of words and morphemes. Wasn't it the case that words are sometimes formed after the syntax, for instance by morphologically merging the features in Fin with V+v?

Search Results

Refine search

Refine search

Actions for selected content:

23 results

Introducing Syntax

Introduction: The Language Machine

Summary

Acknowledgements

7 - Unifying Movement and Agreement

Summary

5 - Agreement and Uninterpretable Features

Summary

Index

Afterword

Summary

1 - Categories and Features

Summary

6 - Movement and Remerge

Summary

3 - Theta Theory (θ-Theory)

4 - Case Theory

Summary

9 - Syntax and Phonology

Summary

Foreword

Summary

Contents

2 - Merge

Summary

Frontmatter

About this Book

Summary

Glossary

References

10 - Syntax and Semantics

Summary

Search Results

Refine search

Refine search

Actions for selected content:

Save Search

23 results

Introducing Syntax

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary