2.1.2 Phrases

2.1.3 Sentences within phrases

2.1.4 Structural positions

Notice that in all these structures the verb occupies the same position inside the second phrase of the sentence. It would not matter how many words or other phrases the first phrase contained, the verb would still be in the same position with respect to its own phrase, and hence grammatical position defined in terms of structure is much more satisfactory than in terms of linear order. Moreover, if we consider the structure of the phrase that the verb appears in, we can identify its position within this phrase more easily than by counting its position in a linear string:

2.1.4 Structural positions

Again, we cannot identify the position of the verb within the phrase in terms of linear order, but in terms of the structure it is clear that the verb always comes directly underneath this phrase, precedes its phrasal complements and may follow adverbial modifiers. As we proceed it will become even clearer that there is a unique structural position that verbs and verbs alone can occupy.

2.1.5 Structural terminology

Now that we have introduced the notion of structure, we need some terms to use to refer to aspects of structures and the way we can represent them. The notion of structure entails that there are elements of a sentence that themselves are made up of other elements and indeed that these other elements may be made up of yet more elements, and so on and so on. The elements that make up a larger part of the structure are called its constituents and the constituents that directly make up a part of structure are called its immediate constituents. Thus, in a phrase such as the following, the verb and its complement phrase are its immediate constituents. Everything inside the complement phrase is a constituent of the whole phrase, though not an immediate constituent:

2.1.5 Structural terminology

This kind of representation of grammatical structure is called a tree diagram, though unlike real trees, grammatical trees tend to grow downwards. The elements that make up the tree, the words and phrases etc. are called nodes and the lines that join the nodes are branches.

2.1.5 Structural terminology

The bracketing in (29) is not entirely equivalent to the tree diagram in (16) as in the tree the nodes have labels that tell us what they represent, phrases or sentences. We can add labels to brackets to make the two representations equivalent. With bracketing, the label is usually placed on the open bracket of the constituent:

2.1.6 Labels

Although we have been labelling phrases with the symbol P, not all phrases are equivalent to each other. This is best seen in terms of the distributions of phrases. Take, for example the two phrases in (16) the postwoman and the doctor. These look very similar, both consisting of a determiner followed by a noun. They also have the same distribution patterns, as shown by the fact that wherever we can put one of them we will also be able to put the other:

2.1.6 Labels

As these phrases have the same distributions, we can assume that they are phrases of the same kind. However, not all phrases distribute in the same way. Consider the phrase on his birthday. This cannot go in the same places as those in (32):

2.1.6 Labels

Clearly this phrase must be different from the previous two. We will see in the next chapter that the identity of a phrase is determined by one of the words it contains. This word is known as the head of the phrase. It will be argued later on in this book that the head of phrases such as the postwoman is the determiner and the head of phrases such as on his birthday is the preposition. Thus, we distinguish between determiner phrases (DPs) and preposition phrases (PPs).

2.1.6 Labels

There are also other phrases associated with the verb (VPs), with adjectives (APs) and indeed with every kind of word category that we have discussed (noun phrases – NPs, inflectional phrases – IPs, CPs and degree adverb phrases – DegPs).

2.1.6 Labels

For now, the main point is that there are different kinds of phrases and these have different positions within the structure of the sentence and hence different distributions. We might therefore represent the sentence in (31) more fully as:

2.1.7 Rules

This fact does not invalidate this kind of rule for linguistic descriptive purposes. As long as there is only a finite number of rules, a legitimate grammar could be formulated even with a very large number of them. However, if human grammars are constructed of a large number of rules the question is raised of how children could ever learn their grammatical systems. This consideration has lead some linguists to assume that what is needed is a far more restricted set of rules. We will introduce the theory of phrase structure that follows this line of thought in the next chapter.

2.3.3 Coordination

Not just words can be coordinated however; we can also coordinate phrases and sentences. As long as the phrases and sentences are sufficiently the same, the result will be a phrase or a sentence which behaves in the same way as its coordinated parts:

2.3.3 Coordination

Again we can turn these observations round to provide a test for structural analyses. If we claim that a certain part of a sentence constitutes a phrase, then to test this claim we could take another similar element and see if the two things can be coordinated. Thus, to go back to the structure proposed in (100), there are three constituents proposed: the subject DP, the VP and the object DP inside the VP. If this is accurate, we should be able to find an element to coordinate with these constituents to form grammatical sentences:

2.3.4 Single-word phrases

3.1.1 Rewrite rules and some terminology

We will start by looking at some general principles that determine the basic structure of phrases and sentences. The perspective we will present claims that these principles are simple because there are a very small number of them that apply to all structures. In fact this theory claims there to be at most three different rules which determine the nature of all structures in a language. These can be stated as follows:

3.1.1 Rewrite rules and some terminology

Recall from chapter 2, rewrite rules which tell us how structures of various kinds decompose into their constituent parts. The rules in (1) are like these, only far more general. The generality is achieved through the use of category variables, X and Y, which stand for any possible category (nouns, verbs, prepositions, determiners, etc.). Thus these rules tell us how phrases in general are structured, not how particular VPs, PPs or DPs are.

3.1.1 Rewrite rules and some terminology

The third rule in (1) introduces a position into the phrase called the adjunct. Given that we have yet to introduce these elements we will put off discussion of this rule until section 1.3. where we will give a fuller account of both adjuncts and the adjunction rule.

3.1.1 Rewrite rules and some terminology

There are several things to note about this structure. First there are two immediate constituents of the X' (pronounced “X bar”): X, which is called the head of the phrase and the complement YP. The complement, which, as its label suggests is a phrase of any possible category, follows the head. This is a fact about English and in other languages the complement may precede the head.

3.1.1 Rewrite rules and some terminology

Whether it precedes or follows the complement, the head is the central element of the phrase and is a word of the same category as the X'. Thus, if the head is an adjective, the X' will be an A' and if the head is a complementiser the X' will be a C'.

3.1.1 Rewrite rules and some terminology

Note that, although these are constituents of different types, they all have a very similar pattern: the head is on the left and the complement is on the right. This is exactly what the X-bar rules were proposed to account for. It is clearly the case that there are cross-categorial generalisations to be made and if constituents were described by the rewrite rules of the kind given in chapter 2, where for each type of constituent there is a specific rule, it would be impossible to capture obvious similarities between phrases.

3.1.1 Rewrite rules and some terminology

Again there are several things to note about this structure. Once more, there are two immediate constituents of the phrase. The specifier, a phrase of any category, precedes the X', the constituent just discussed containing the head and the complement. Again the ordering of these two constituents is language dependent: specifiers precede X's in English, but this is not necessarily so in all languages. Specifiers are a little more difficult to exemplify than complements due to complications that we have yet to discuss. However, the following are fairly straightforward cases:

3.1.1 Rewrite rules and some terminology

Note that the X' and the phrase share the same categorial status (X) and so if X' is P' XP will be PP, etc. As X' is the same category as the head, it follows that the whole phrase will be of the same category as the head. In this way, the head of the phrase determines the phrase’s category.

3.1.1 Rewrite rules and some terminology

The property of sharing category between the head, the X' and the phrase is called projection. We say that the head projects its categorial status to the X' and ultimately to the XP. If we put the two parts of the structure together, we can more clearly see how projection works:

3.1.1 Rewrite rules and some terminology

The line of projection proceeds from the head, via the X' to the phrase thus ensuring that phrases and heads match.

3.1.1 Rewrite rules and some terminology

The meaning of the ‘bar’ can be seen in terms of the notion of projection. We can imagine a phrase as a three-floored building, with a ground floor, a first floor and a top floor. On the ground floor we have the head, which is not built on top of anything – it is an unprojected element. Often heads are called zero level projections, to indicate that they are not projected from anything. This can be represented as X⁰.

3.1.1 Rewrite rules and some terminology

On the top floor we have the phrase, XP. This is the highest level projected from the head and hence it is called the maximal projection. Another way of representing the maximal projection is X'', an X with two bars (pronounced ‘X double bar’), with the bars again representing the projection level. It seems that all phrases project to two levels and so we will not entertain the possibility of X''', or X'''', etc. Typically we will maintain the custom of representing the maximal projection as XP.

3.1.2 Endocentricity

An obvious consequence of the notion of projection is that we will never get a phrase of one category with a head of another. While this might seem a slightly perverse situation to want to prevent in the first place (why would verb phrases be headed by anything other than a verb?), it is certainly a logical possibility that there could be phrases of category X which do not contain a word of category X. For example the traditional view that preposition phrases can function adverbially could be captured under the following assumption:

3.1.2 Endocentricity

In other words, a preposition phrase which behaves as an adverbial phrase is an adverb phrase headed by the preposition. Clearly this is something that would not be allowed by the X-bar rules in (1). Evidence favours the X-bar perspective and there is no reason to believe that just because something functions adverbially it is categorially the same as an adverb. For example, even when PPs are used adverbially, they still have different distributions to AP:

3.1.2 Endocentricity

As we see in (9), a PP modifier of a verb must follow it, while an AP modifier may precede or follow it, even if the two modifiers have virtually the same interpretation. Thus a phrase headed by a preposition has a different categorial status to one headed by an adverb, supporting the X-bar claim that phrases have heads of the appropriate kind.

3.1.2 Endocentricity

Moreover, the X-bar rules in (1) rule out another possibility if we assume that these are the only rules determining structure. While it might not make much sense to have a phrase with a head of a different category, the idea of a phrase that simply lacks a head is not so absurd. There is a traditional distinction made between endocentric and exocentric language elements. An endocentric phrase gets its properties from an element that it contains and hence this element can function by itself as the whole phrase. For example:

3.1.2 Endocentricity

An exocentric phrase on the other hand contains no element that can have the same function as the whole phrase and so appears to have properties that are independent from the elements it contains. A standard example is:

3.1.2 Endocentricity

The issue is rather complex. The traditional view mixes category and function in a way that is perhaps not helpful. The point is, however, that the X-bar rules in (1) claim that, categorially, all phrases are endocentric: in other words, all phrases have heads which determine their categorial nature.

3.1.2 Endocentricity

As we see from (15), a by phrase containing a reflexive is interpreted to mean ‘unaccompanied’. In (15a), the reflexive refers to Pete and so it means that he was unaccompanied in eating the pie. In contrast, in (15b) the reflexive refers to the pie and so it means that the pie was unaccompanied (by ice cream for example) when Pete ate it. (16a) is grammatical even though there is no apparent antecedent for the reflexive. It is not surprising that the reflexive should be yourself however, as, as we have said, the understood subject of an imperative is you. Yet we cannot simply say that the antecedent ‘being understood’ is enough to satisfy the requirements of the reflexive as (16c) is ungrammatical. In this case the object is absent, though it is clearly understood that something was eaten in (16b). But this understanding is not enough to license the use of the reflexive in this case. So we conclude that the missing subject in (16a) is different from the missing object in (16c) and in particular that the missing subject has a more definite presence than the missing object. This would be so if the missing subject were present as an unpronounced grammatical entity while the missing object is absent grammatically and present only at the semantic level. In conclusion then, while imperatives might look like VP clauses which lack subjects, they are in fact full clauses with unpronounced subjects.

3.1.3 Heads and Complements

But if the structural rules of the grammar are themselves so general as to not make reference to categories how do categorial features come to be in structures? One would have thought that if all the grammar is constructed from are rules that tell us how phrases are shaped in general, then there should only be one kind of phrase: an XP.

3.1.3 Heads and Complements

To see how categorial information gets into structure we must look more closely at heads and the notion of projection. We have seen how heads project their properties to the X' and thence to XP, the question we must ask therefore is where do heads get their properties from? The main point to realise is that the head is a word position and words are inserted into head positions from the lexicon. In chapter 1 we spent quite some time reviewing the lexical properties of words, including their categorial properties. These, we concluded, are specified for every lexical item in terms of categorial features ([±F, ±N, ±V]). If we now propose that a head’s categorial features are projected from the lexical element that occupies the head position we can see that phrases of different categories are the result of different lexical elements being inserted into head positions.

3.1.3 Heads and Complements

In this way we can see that categorial features are actually projected into structures from the lexicon. This makes a lot of sense given that categorial properties are to a large extent idiosyncratic to the words involved in an expression and are not easily predicted without knowing what words a sentence is constructed from. Things which are predictable, such as that all phrases have heads which may be flanked by a specifier to its left and a complement to its right, are what are expressed by the X-bar rules. Thus we have a major split between idiosyncratic properties, which rightly belong in the lexicon, and general and predictable properties which rightly belong in the grammar.

3.1.3 Heads and Complements

The structure in (21) is still incomplete however and we must now consider how to complete it. Let us concentrate firstly on the complement position. At the moment this is expressed by the general phrase symbol YP. This tells us that only a phrasal element can sit here, but it does not tell us what category that phrase must be. Yet, it is clear that we cannot insert a complement of just any category into this position:

3.1.3 Heads and Complements

As we know that the complement must be prepositional, we also know by the general principles of X-bar theory that only a preposition could be inserted into the head position of this phrase and the lexical properties of this head will, in turn, impose restrictions on what can appear in its complement position:

3.1.3 Heads and Complements

In this structure, off is inserted into the head position of the PP complement of the verb and as this preposition selects for a DP complement (as most of them do), only a determiner could be inserted into the head position of this phrase. The determiner would then impose restrictions on its complement, ensuring this to be an NP and hence only a noun could be inserted into the head of the determiner’s complement. Obviously, this could continue indefinitely, but in this case the process stops at this point as the noun subcategorises for no complement.

3.1.5 Adjuncts

This is different from the previous two rules in a number of ways. First, the previous rules specified the possible constituents of the various specific projections of the head: complements are immediate constituents of X' and specifiers are immediate constituents of XP. The adjunction rule in (30) is more general as it states the possible constituents of an Xⁿ, that is, an X with any number of bars. In other words, Xⁿ stands for XP (=X''), X' or X (=X⁰). The adjunct itself is defined either as a word (Y) or as a phrase (YP) and we will see that which of these is relevant depends on the status of Xⁿ: if Xⁿ is a word, then the adjunct is a word, if not then the adjunct is a phrase.

3.1.5.1 Adjunction to X-bar

The bracketed elements in (32a) and (b) have the same distribution and hence we can conclude they have the same categorial status. As this phrase in (32b) contains only a noun, we conclude that it is an NP. In (32c) however, the phrase following the determiner contains only an adjective and is ungrammatical. This clearly has a different distribution to the other two phrases, indicating that the adjective in (32a) is not the head of this phrase.

3.1.5.2 Adjunction to phrase

3.1.5.3 Adjunction to head

An interesting point to note is that the adjunct to a head is always a head itself, which differs from the previous cases of adjunction we looked at above. Adjuncts adjoined to X' or XP are always phrases. It has been suggested that this is due to a restriction on adjunction such that only like elements can adjoin: heads to heads, phrases to phrases. If this is true, then X' adjunction should not be possible as the adjunct differs in its X-bar status to the X', being a phrase. We will not accept this point of view however and assume that while only heads can adjoin to heads, phrases can adjoin to any constituent larger than a head.

3.1.6 Summary

Before moving on to look at other aspects of syntactic processes, let us consolidate what we have said in this section. X-bar theory is a theory of basic structure comprising of just three rules. These rules are generally applicable to all structures and substructures, no matter what their category: they are category neutral. The categorial status of a specific structure depends on the lexical elements it contains, in particular one word acts as the head of each phrase and this determines the category of the phrase by projecting its own categorial properties, established in the lexicon, to the X' node above it and ultimately to the XP.

3.1.6 Summary

The three X-bar rules introduce three elements besides the head. The complement is introduced as the sister of the head. It always follows the head and is restricted by the head’s subcategorisation requirements. Thus, if a head selects for a PP complement, the complement must be a PP. The specifier is introduced as the sister to X' and daughter of XP. Specifiers precede the head and are restricted to one per phrase. The last element of the phrase, the adjunct, can be introduced at any X-bar level: X, X' and XP. This element expands what it is adjoined to into another element of the same type. Therefore the process is recursive and in principle any number of adjuncts can be added to a structure.

3.2 Theoretical Aspects of Movement

The obvious questions to ask are: why is the object sitting at the front of the sentence in (46)?; and how is the interrogative pronoun interpreted as an object when it is not sitting in an object position? As to the first question the obvious answer is that it has something to do with the interrogative nature of the clause: the clause is a question and interrogative clauses of this kind start with an interrogative phrase such as who.

3.2.1 Move α

Here the main verb wasted takes its argument in the specifier position of its own phrase. This phrase is in turn the complement of the auxiliary verb was. The argument moves from the specifier of the lower verb to that of the higher one.

4.1 Why the Noun is not the Head of the DP

This is not a good argument, however, as it is not wise to conclude about the syntactic properties of an element on the basis of its semantic properties. There are many elements which might be considered to be the syntactic head of a phrase which are not the semantically most important word. For example, consider the following:

4.1 Why the Noun is not the Head of the DP

Semantically, the noun tea is the most important element in this phrase: it refers to something which can be described as tea and not a cup. When one drinks a cup of tea, it is the tea that gets drunk, not the cup! Yet, syntactically it seems that cup should be considered as the head and the phrase containing tea as its complement. This provides us with a straightforward structure:

4.1 Why the Noun is not the Head of the DP

If, on the other hand, we wanted to claim that the noun tea is the syntactic head of the phrase we would have difficulty fitting in the preposition and the other noun:

4.1 Why the Noun is not the Head of the DP

Another case where it might be argued that the syntactic head of a phrase is not the most important semantic element within it concerns preposition phrases:

4.1 Why the Noun is not the Head of the DP

In these cases, as in those above, the preposition does contribute something to the meaning of the phrase, though it is not clear that this should be seen as the most important aspect of the meaning of the whole phrase. Indeed London and tunnel seem to contribute just as important, if not more important information. However, it would not make sense to claim that the nouns are the heads of these phrases as they are clearly not NPs, not having the distribution of NPs:

4.1 Why the Noun is not the Head of the DP

A second observation that might support the assumption that the noun and not the determiner is the head of the phrase is the fact that the noun contributes features which play a role in interpreting the meaning of the whole phrase:

4.1 Why the Noun is not the Head of the DP

In (8a) the whole phrase is considered to be singular and in (8b) the phrase is plural, as can be observed from facts concerning verb agreement:

4.1 Why the Noun is not the Head of the DP

In these cases both the nouns and the determiners are marked for number and so it is difficult to say where the number feature of the whole phrase is projected from. Indeed, even in those cases such as (8a) and (8b) where it looks as though the number is projected from the noun, we could argue that the determiner the is ambiguously marked for singular or plural and, like the other determiners, when it is singular it can only accompany a singular noun and when it is plural it can only accompany a plural noun. The issue therefore rests on which we take to be the head: the determiner or the noun. For this reason, we cannot use these observations to argue in favour of one or the other having head status but we must look elsewhere to resolve the issue.

4.1 Why the Noun is not the Head of the DP

Determiners, on the other hand, are not recursive and have a very fixed position at the beginning of the phrase:

4.1 Why the Noun is not the Head of the DP

The assumption that the determiner is the head of the phrase, on the other hand, captures its position perfectly: it precedes the noun because the noun heads its complement and heads precede their complements in English. Comparing the two options, then, it seems that the one in which the determiner is the head is the more straightforward:

4.1 Why the Noun is not the Head of the DP

Another problem that arises if we assume that determiners are not heads of phrases is that they do make a contribution to the whole phrase. In chapter 1 we spent some time discussing properties of determiners (section 3.5.2), pointing out that a major contribution determiners make to the phrases that contain them is the definiteness–indefiniteness distinction: