This article describes the syntax of clauses in the English language, chiefly in Modern English. A clause is often said to be the smallest grammatical unit that can express a complete proposition. But this semantic idea of a clause leaves out much of English clause syntax. For example, clauses can be questions, but questions are not propositions. A syntactic description of an English clause is that it is a subject and a verb. But this too fails, as a clause need not have a subject, as with the imperative, and, in many theories, an English clause may be verbless. The idea of what qualifies varies between theories and has changed over time.
History of the concept The earliest use of the word clause in Middle English is non-technical and similar to the current everyday meaning of phrase: "A sentence or clause, a brief statement, a short passage, a short text or quotation; in a ~, briefly, in short; (b) a written message or letter; a story; a long passage in an author's source."
The first English grammar, Pamphlet for Grammar by William Bullokar, was published in 1586 and briefly mentions clause once, without explaining the concept.
A technical meaning is evident from at least 1865, when Walter Scott Dalgleish describe a clause as "a term of a sentence containing a predicate within itself; as... a man who is wise."
In the early days of generative grammar, new conceptions of the clause were emerging. Paul Postal and Noam Chomsky argued that every verb phrase had a subject, even if none was expressed, (though Joan Bresnan and Michael Brame disagreed). As a result, every verb phrase (VP) was thought to head a clause.
The idea of verbless clauses was perhaps introduced by James McCawley in the early 1980s with examples like the underlined part of with John in jail... meaning "John is in jail".
Types of clause in Modern English Clauses can be classified as independent (main clauses) and dependent (subordinate clauses). An orthogonal way of classifying clauses is by the speech act they are typically associated with. This results in declarative (making a statement), interrogative (asking a question), exclamative (exclaiming), and imperative (giving an order) clauses, each with its distinctive syntactic features. Declarative and interrogative clauses may be independent or dependent, but imperative clauses are only independent.
Dependent clauses have other cross-cutting types. These include relative and comparative clauses; and participial and infinitival clauses.
Finally, there are verbless clauses.
Examples +English Clause typesIndependentDependent (underlined)DeclarativeThis is a tree.I think (that) this is a tree.InterrogativeIs this a tree? What is this?I wonder whether / if this is a tree. I wonder what this is.ImperativeBe still. Stop.ExclamativeWhat a tall tree that is!Did you see what a tall tree it is?RelativeThis is the tree (that) I saw. The tree is in my yard, where I planted it.
This is a tree to behold.ComparativeThis tree is taller than that one is.InfinitivalI can see it. I want to see it.ParticipialI'm looking at the tree. I'm shaded by the tree.VerblessWhat a great thing to do!These clauses each include a verb, but the verb is part of a subordinate clause. How odd that nobody noticed!
Out of my way!He stood there quietly, his hands in his pockets. Whether a fact or not, it is commonly believed.
Whatever the reason, it's annoying.
He didn't love her as much as she him.
The bramble shook as if alive.
Independent clause types
Declarative By far, the most common type of English clause is the independent declarative. The typical form of such clauses consist of two constituents, a subject and a head verb phrase (VP) in that order, with the subject corresponding to the predicand and the head VP corresponding to the predicate. For example, the clause Jo did it has the subject noun phrase Jo followed by the head VP did it. Declarative clauses are associated with the speech act of making a statement.
The following diagram shows the syntactic structure of the clause this is a tree. The clause has a subject noun phrase (Subj: NP) this and a head verb phrase (Head: VP). The VP has a head verb is and a predicative complement NP (PredComp: NP) a tree.
Information packaging constructions can result in the addition of other constituents and various constituent orders. For example, the it-cleft construction has it as a dummy subject, followed by a head VP containing a form of the verb be + a complement corresponding to the predicand + a relative clause whose head corresponds to the predicate. So, the example above as an it-cleft is It was Jo who did it.
V2 Some declarative clauses follow V2 order, which is to say the first verb appears as the second constituent, even if the subject is not the first constituent. An example would be Never did I say such a thing, where never is the first constituent and did is the verb in V2 position. Addition of a tensed form of the auxiliary verb do is called do support.
Interrogative There are two main types of independent interrogative clauses: open and closed. These are most associated with asking questions, but they can be used for other speech acts such as giving advice, making requests, etc.
Open interrogatives include an interrogative word, which, in most cases either is the subject (e.g., Who went to the shop?) or comes before an auxiliary verb + the subject. This is seen in What can you buy there? where what is the interrogative word, can is the auxiliary, and you is the subject. In such cases, the interrogative word is said to be fronted, or it may be part of a fronted constituent, as in which shop did you go to? When no auxiliary verb is present then do support is required.
The interrogative word can also appear in the non fronted position, so that the example above could be You can buy what there? where what is an object in the VP. When it is fronted, many modern theories of grammar posit a gap in the non-fronted position: What can you buy __ there? This is a kind of discontinuity.
The following diagram shows the syntactic structure of the clause What can you buy there? The clause has a fronted noun phrase (Front: NP) what, which is co-indexed to the object gap in a lower VP.
Closed interrogative clauses can be further subdivided as polar or alternative. A polar interrogative is one to which the expected response is yes or no. For example, Do you like sweets? is a polar interrogative and another case of do support. An alternative interrogative is one asking for a choice among two or more alternatives, as in Would you like coffee or tea? In both types of closed interrogatives, an auxiliary verb is fronted. That is to say, it comes before the subject. In the example above, would is the fronted auxiliary verb and you is the subject.
Another minor clause type is the interrogative tag. A tag is appended to a statement and includes only an auxiliary verb and a pronoun: you did it, didn't you?
Imperative In most imperative clauses the subject is absent: Eat your dinner! However imperative clauses may include the subject for emphasis: You eat your dinner! In either case, the predicand is understood to be the person being addressed. There is also an imperative construction with let and the first person plural, as in let's go. An example like let them go is still understood as having a second-person predicand.
Imperatives are closely associated with the speak acts of commands and other directives.
The verb in an imperative clause is in the base form, such as eat, write, be, etc. Negative imperatives uses do-support, even if the verb is be; see below.
Exclamative Exclamative clauses start with either the adjective what or the adverb how and are typically associated with exclamations. As with open interrogatives, the what or how phrase is fronted unless – in the case of what – it's the subject.
What great students you have! (subject) What a nice thing you did. (object gap) How kind you are. (predicative complement gap) The following diagram shows the syntactic structure of the clause How kind you are. The clause has a fronted adjective phrase (Front: AdjP) how kind, which is co-indexed to the predicative complement gap (PredComp: gap) in the VP.
Dependent clause types Clauses can be nested within each other, sometimes up to several levels. These clauses within clauses are said to be dependent. For example, the sentence I know the woman who says ! contains the following dependent clauses: a non-finite clause (drinking beer) within a content clause (she saw your son drinking beer) within a relative clause (who says she saw your son drinking beer). These are all within the independent declarative clause (the whole sentence).
As the example above shows, a dependent clause may be finite (based on a finite verb, as independent clauses are), or non-finite (based on a verb in the form of an infinitive or participle). Particular types of dependent clause include relative clauses (also called "adjective clauses"), content clauses (traditionally called "noun clauses" and also known as "complement clauses") and comparative clauses, each with its own characteristic syntax.
Traditional English grammar also includes adverbial clauses, but since at least 1924, when Jespersen published The philosophy of grammar, many linguists have taken these to be prepositions with content clause complements.
Relative clauses
Syntactically, relative clauses (also called "adjective clauses") typically contain a gap (as explained above in interrogative clauses). Semantically, they contain an anaphoric relation to an element in a larger clause, typically to a noun. There are two main relative clause types: wh- relatives and non-wh- relatives, the latter of which can be further subdivided into that and bare types. +Examples of relative clause typesExampleType1.This is the house which Jack built.wh-2a.This is the house that Jack built.non wh-: that2b.This is the house Jack built.non wh-: bare The semantic relation can be seen most easily in (1) above. This clause has a gap in the VP headed by built, where an object would usually appear. For the purposes of illustration, the gap is replaced by it in the following diagram.
This shows an anaphoric relation inside the relative clause between the gap (filled by the resumptive pronoun it), and the fronted relative pronoun which. It shows a second anaphoric relation between the relative pronoun and the noun in the main clause the house. This means "this is the house" and also "Jack built the house". In a wh- relative, when the related item in the relative clause is the subject of the relative, there is no gap, so there is only the anaphoric relation between the relative pronoun and an element in the main clause (e.g., Jack, who built the house, is a good chap.)
Non-wh- relatives Non-wh- relative clauses are of two types: that clauses and bare clauses. In most cases, either one is possible, as shown in (2) above, but when the relative item is the subject of the relative clause, there is a gap in the subject position, and bare relatives are not possible (e.g., these are the folks that __ have been helping, but not *these are the folks __ have been helping.)
Traditional grammar calls that a relative pronoun, like who above, but modern grammars consider it to be a complementizer, not a pronoun.
Non-wh- relative clauses are not typically possible with supplementary relatives. (See the main article on English relative clauses for the distinction between integrated and supplementary relatives.)
Wh- relatives Wh- relative clauses include a relative word, a pronoun who or which, a preposition when or where, and adverb how, or an adjective, also how. This is fronted, leaving a gap, unless it is the subject or part of the subject.
Comparative clauses