glossary online

Return to home page

Grammatical terminology recommended for use in schools

© Linguistics Association of Great Britain, 2015

Explanatory note for school teachers and publishers

Schools need a unified terminology for grammar just as they do for any other subject, and for the same reasons — to provide consistency between teachers within a single school, and to provide consistency across schools (and between school and university). Consistency within a school is particularly important if English teachers and foreign-language teachers are to support each other as they should.

The aim of this glossary is to provide a set of grammatical terms which could be adopted by schools and textbook writers. It has been written and agreed by grammarians in UK universities, with the twin aims of providing the best possible combination of accessibility to school teachers and acceptability to grammarians. At the university level, grammar is a very active research area within linguistics and has seen enormous growth and developments since the 1950s; not surprisingly, this activity has produced disputes and divisions among grammarians, so the glossary is an exercise in compromise: compromise between the needs of schools (including the terminology already recommended in the National Curriculum) and those of universities, and compromise among proponents of different approaches to grammar.

The terms selected are mostly relevant to English grammar, but many of the entries for these terms mention similarities and differences between English and commonly-taught foreign languages. In addition, there are a few entries for terms which are relevant only to foreign-language teaching.

The glossary is simply a reference tool, from which teachers and textbook writers can select according to their professional judgement of pedagogical needs. No suggestion is intended that every teacher should know every term, less still that every school leaver should.

As in any other technical area, grammatical terms are tightly integrated into a complex network and are therefore hard to present in isolation, and even basic terms have to be defined in relation to a range of other terms. The electronic medium of the glossary allows hyperlinks to reveal these interconnections, and users can follow links when needed — but the hope is that most hyperlinks will become redundant for most users.

Explanatory note for grammarians

Grammarians should be aware of the following controversial assumptions that are made in the glossary:

  • Phrases: Phrase structure is assumed (rather than dependency structure), but phrases are only recognised when they consist of more than one word. This produces a lack of generality because of the repetition of ‘X or X-phrase’, but it avoids the pedagogical problems of unary branching. But clauses are exceptional, so Hurry! is both a verb and an imperative clause.

  • Phrase classes are allowed to diverge from those of their head word; for instance, an infinitival clause is headed by a base-form verb (not by an infinitive).

  • Noun phrases and determiners: Noun phrases are always headed by nouns, not by determiners. The function ‘specifier’ (inside the NP) is contrasted with the word-class ‘determiner’ and the category ‘genitive’.

  • Pronouns and determiners: Pronouns are treated as a subclass of noun, and ‘determiner’ is recognised as a top-level word class. Some lexemes belong to both classes.

  • Inflectional forms: these are only recognised when there is some morphological evidence — ie. total syncretism is not allowed, so (for example) ‘imperative’ and ‘infinitive’ cannot be distinct inflectional forms. Instead, they are distinguished in the glossary at the level of the clause.

  • Auxiliaries and VPs: Under ‘clause’, the glossary notes three analyses of auxiliary+verb sequences without committing to any. It recommends avoiding the term verb phrase.

  • Complement: the glossary recognises ‘complement’ as a general category subsuming ‘object’ and ‘subject/object complement’.

  • Tense and aspect: the glossary recognises periphrastic tenses as well as the simple inflected ones, while recognising ‘aspect’ as a possible name for progressive and maybe perfect; but it argues against a periphrastic future tense.

  • Conjunctions: divided between ‘coordinator’ and ‘subordinator’.

  • Terminology: the glossary prefers:

    • progressive to continuous

    • restrictive to defining

    • genitive to possessive

    • modifier to adjunct

    • preposition phrase to prepositional phrase

    • numeral to number

Other reference works

This glossary duplicates information which is easily available in a number of published books, and in general its definitions are compatible with those found in these works (which, in turn, are generally in agreement with each other). What is distinctive about this glossary is its relative brevity, due to its focus on education, and its electronic format — and, of course, its free availability. Those who want a more comprehensive glossary will find any of the following both reliable and accessible:

  • Bas Aarts, Sylvia Chalker and Edmund Weiner: The Oxford Dictionary of English Grammar. Second edition. (Oxford University Press 2014)
  • David Crystal: A Dictionary of Linguistics and Phonetics. (Blackwell 1980 and later editions)
  • Peter Matthews: Concise Oxford Dictionary of Linguistics. (Oxford University Press 1997)
  • Larry Trask: A Dictionary of Grammatical Terms in Linguistics. (Routledge 1993)

The glossary also tries to reflect a consensus view of the terminology found in the main recently-published grammars of English:

  • A Comprehensive Grammar of the English Language (Quirk, Greenbaum, Leech, Svartvik. Longman, 1985)
  • Longman Grammar of Spoken and Written English (Biber, Johansson, Leech, Conrad and Finegan. Longman, 1999)
  • The Cambridge Grammar of the English Language (Huddleston and Pullum. Cambridge University Press, 2002)
  • Cambridge Grammar of English (Carter and McCarthy. Cambridge University Press, 2006)
  • Oxford Modern English Grammar (Aarts. Oxford University Press, 2011)


abbreviation. An abbreviation is a shortened way of writing a word or group of words; it is often the result of clipping. For example: Co. (Company), approx. (approximately), PR (public relations). A few common abbreviations are of Latin terms (for example: e.g. = exempli gratia = for example). Names of organisations are often abbreviated using the initial letters of each word (e.g. the NHS (National Health Service)). Some such abbreviations (e.g. NATO, FIFA and UNESCO) are pronounced like ordinary words and are called ‘acronyms’. See also contraction.

abstract noun. Nouns such as beauty, time, hour and grammar are often classified as abstract in contrast with concrete nouns such as dog, tree, stone and person. However, this is not a grammatical distinction, because it does not affect the grammar of the words concerned – abstract and concrete nouns follow exactly the same rules. It is only a matter of meaning, so the same noun may sometimes have a concrete meaning, and at other times an abstract one, without affecting its grammar; for instance grammar may name either a subject of study (She got tired of studying grammar) or a concrete object made of paper and cardboard (She threw her grammar at the teacher.)

accusative. See case.

acronym. See abbreviation

active. See voice.

addressee. The addressee is ‘you’, i.e. the person (or people) who are intended to receive the message — the intended listener or the intended reader.

adjective. E.g. big, extensive, vertical. A typical adjective can be used in two different ways:
• either before a noun (e.g. big box), acting as the noun’s modifier. This use is called ‘attributive’.
• or after the verb be (e.g. is big), or other linking verbs such as seem, where it functions as the verb’s subject complement (e.g. seems nice) Because subject complements are part of what is sometimes called ‘the predicate’, this use is called ‘predicative’.

Adjectives are relatively easy to identify by means of grammatical characteristics such as these, but meaning is an unreliable guide. Adjectives are sometimes called ‘describing words’ because they often pick out single characteristics of people and things such as size or colour. This definition is unhelpful because it doesn’t distinguish adjectives from other word classes such as verbs, nouns and adverbs which can do the same. For example, the verb shimmered describes the water in The water shimmered; the noun idiots describes them in They are idiots; and the adverb softly describes the speaking in She spoke softly. Many adjectives are gradable, and can be modified by very (very big, very expensive) which can only modify adjectives, adverbs and some quantifiers. Short gradable adjectives such as big and short also have comparative and superlative forms: bigger, biggest; shorter, shortest, and the same effect can be achieved with longer adjectives by combining them with more and most: more intelligent, most intelligent.

adjective phrase. E.g. very nice, good enough to sell. An adjective phrase is a phrase whose head is an adjective.

adverb. E.g. quickly, fortunately, soon, almost, very. Adverbs are often used as modifiers of a verb (hence the name ad-verb) to add more details to its meaning, such as its manner, time, or place (e.g. She arrived quickly. She will leave soon). In this use, they function as adverbials.
Many adverbs are formed by adding -ly to an adjective (quick-ly, fortunate-ly) so these adverbs are easy to recognise, and help to identify others which may replace them. For instance, often and almost must be adverbs because of their similarities in both syntax and meaning to the more easily recognised adverbs frequently and nearly.

Verbs are not the only words to which adverbs may be added. They may also be used as degree modifiers with adjectives and other adverbs:

  • adjectives (nearly impossible, extremely good)
  • other adverbs (almost impossibly difficult, She spoke quite clearly).
    In addition, adverbs occasionally even modify determiners or prepositions (She ate almost every cake, The rug reached almost to the wall). adverb phrase. E.g. very carefully, so recently that I can still remember it. An adverb phrase is a phrase whose head is an adverb.

adverbial. In Recently, I met my neighbour in the street, both recently and in the street are adverbials — parts of the clause which modify the verb. Like the more general term modifier, the term adverbial is the name of a grammatical function, not a word class, though it is named after adverbs, one of the word classes that may be used as adverbials (e.g. recently). But adverbs are not the only possibilities, or even the most common. Other possibilities are a preposition phrase (in her garden), a noun or noun phrase (I saw her this morning) or a subordinate clause (She was there when I arrived). The term ‘adverbial’ is not usually applied to modifiers of other word classes; so unexpectedly is an adverbial in She arrived unexpectedly, but not in She did unexpectedly well (where it modifies well, not the verb did).

  • Some adverbials apply to the meaning of the entire clause (Fortunately, she was unharmed. She was actually joking. In fact, it only rained a little).

adverbial clause. See subordinate clause.

affirmative. A clause may be classified as either affirmative or negative; for example, It is raining and Somebody called for you are affirmative, but can be made negative by adding not or nobody: It is not raining and Nobody called for you. The term ‘positive’ is often used as a synonym of affirmative, and the contrast between affirmative (or positive) and negative is called ‘polarity’.

affix, affixation. An affix is a morpheme which cannot itself be an entire word, and is always attached to a base. An affix can be
• a prefix, added before the base (e.g. intolerant, dislike)
• a suffix, added after the base (e.g. kindness, playing).

It may be used to signal either inflection (e.g. playing) or derivation (e.g. player)

agent. Agent is the name of a semantic role. If a verb denotes an action, the person or thing that carries out the action is the agent. The agent is the ‘do-er’ of the action. For instance, in Mary caught the ball, Mary (the person, not the word) is the agent, and similarly, the ball is called the patient. In a non-action clause such as Mary was happy or Mary caught flu, there is no agent, so the subject of a clause cannot be defined as the agent (or do-er or performer of the action, as it quite often was in older school textbooks). For the same reason, the optional by phrase in a passive clause should be called simply ‘by phrase’ rather than ‘agent phrase’, and passives without a by phrase should be called ‘short passives’ rather than ‘agentless passives’.

agree, agreement. In some cases a verb has different forms with different subjects, so the verb and subject are said to ‘agree’. In Standard English, this happens with all present-tense verbs (except modal auxiliaries), which have –s when the subject is singular and third person but not otherwise:
She likes — they like — I like
does – John and Mary do — I do

It also happens with the verb be in the past tense: she was – they were.

Note that in English (unlike many other languages) singular collective nouns (eg team, family, government) can take a singular or plural verb form, according to whether the people concerned are considered as a group (singular) or as individuals (plural). For example: The team (= it) is a big one. The team (= they) are all small.
There are also a few cases where a determiner must agree with a noun according to whether it is singular or plural. For example:
this house    these houses
Some languages have very rich and complex agreement systems; for example, in German determiners and attributive adjectives agree with the head noun:
der junge Mann wohnt hier. ‘The young man lives here’
die jungen Männer wohnen hier. ‘The young men live here’.

alternative interrogative. See clause type.

ambiguous, ambiguity. A constituent which has more than one possible interpretation is ambiguous. This sometimes arises from unclear grammatical relationships. For example, in the headline: PENSIONER FIGHTS OFF MAN WITH GUN, it is not specified whether the man or the pensioner had the gun. Both interpretations are possible, and either makes sense. Ambiguity is often a source of humour.

anaphora, anaphoric, anaphor. Anaphora is a cohesive device which links one constituent (the ‘anaphor’) to another, its antecedent. For example, in Jill hurt herself, the reflexive pronoun herself relates anaphorically to Jill so herself refers to the same person as Jill; more generally, in any sentence of the form X hurt herself, herself and X refer to the same person.

Similarly, the personal pronoun she relates anaphorically to Emily in I saw Emily yesterday. She told me that she had changed jobs. As this example shows, anaphora may link anaphors and antecedents that are in different sentences. This linkage always involves the meanings, and typically the anaphor and antecedent have the same referent, as in the above example. However anaphora may also relate the anaphor to something implicit in the antecedent, such as the time of the event concerned; for example, then links to the time of the party (implicit in the antecedent had) in We had a lovely party with lots of fun and food. Then we all went to bed. Another possibility is for the anaphor to share the same general category of meaning as its antecedent, rather than the individual referent; for example, the common noun one is interpreted as meaning ‘newspaper’ (rather than some particular newspaper) in I read a French newspaper yesterday and a Spanish one today.

Although anaphora generally works ‘backwards’, i.e. by linking back to an earlier word, it has a special case called ‘cataphora’ in which the anaphor stands before the antecedent. For example, alongside Alan found a marble in his pocket, with anaphoric his, we also find In his pocket, Alan found a marble, where his refers to Alan. Most potentially anaphoric elements also allow ‘exophora’, in which their referent is in the extra-linguistic situation (e.g. Take a look at that, then!)

Anaphora is possible not only for pronouns but also for members of other word classes. Words may be either inherently anaphoric, like pronouns, or anaphoric by ellipsis. Inherently anaphoric words include the following:

  • adjectives, e.g. I prefer the former alternative.
  • adverb, e.g. Meanwhile, let’s have a cup of tea.
  • common noun, e.g. The big ones are nice and ripe.
  • verb, e.g. She may do.)

Anaphora is sometimes described in terms of one constituent ‘referring’ to another, but this is confusing given the established meaning of the term refer in which a word refers to a person or thing, not another word. (The term itself is simply the Greek equivalent of the Latin referre, where ana- and re- mean ‘back’ and phora and ferre mean ‘carry’ — compare their English cognate bear.)

antecedent. Any anaphoric element has an antecedent, the constituent to which it is linked by anaphora. For example, in I asked Mary to help me, but she wouldn’t do it, the words she and it relate anaphorically to their antecedents, Mary and help me; so Mary is the antecedent of she, and help me is the antecedent of it. Similarly, the antecedent of the understood subject in to help me is Mary, and that of a relative pronoun is the constituent (generally a noun) that the relative clause modifies; for example, in people who live in London, the antecedent of who is people.

antonyms, antonymy. Two words are antonyms if their meanings are opposites: hot – cold; light – dark; light — heavy.
A word may have more than one word as an antonym: cold — hot/warm; big — small/tiny/little

apostrophe. An apostrophe is a punctuation mark — a raised comma, as in John’s — used to indicate either omitted letters or a genitive:

  • Omitted letters. We use an apostrophe for the omitted letter(s) when a verb is contracted. For example:
    I’m (I am)
    who’s (who is/has)
    he’d (he had/would)
    she’ll (she will)

In contracted negative forms, not is contracted to n’t and joined to the verb: isn’t, didn’t, couldn’t etc. In some cases, the verb changes its form when combined with n’t: won’t, shan’t, don’t, can’t.
In formal written style, it is more usual to use the full form.
There are a few other cases where an apostrophe is used to indicate letters that are in some sense ‘omitted’ in words other than verbs, eg let’s (= let us), o’clock (= of the clock).

  • Genitive. We also use an apostrophe for the genitive form:
    my mother’s car
    Joe and Fiona’s house

James’s ambition
a week’s holiday
my parents’ car
the children’s clothes

Note that the genitive pronouns yours, his, hers, ours, theirs, and its are not written with an apostrophe.

apposition. When two words or phrases are in apposition to one another, they are simply put next to each other (‘apposed’) so that they can each contribute in different ways to defining the same referent. For instance, in the sentence Our friend Jane came in, Jane is in apposition to our friend because they both refer to the same person by defining different characteristics — her relation to us, and her name.

article. A/an and the are articles. A (an before a vowel sound) is the indefinite article; the is the definite article. Articles are a type of determiner.

aspect. The difference between the progressive I was playing football at five o’clock and I played football at five o’clock is usually called ‘aspect’ rather than tense because it concerns the way in which the event is viewed rather than its position in time, before or after the present moment. The  progressive serves primarily to present the situation as being in progress and limited in duration.

The classification of the perfect as a tense or an aspect is more complicated, as explained in the entry for ‘perfect’.

As explained under ‘tense‘, these two contrasts combine freely with the contrast of tense to define eight distinct ‘tense-aspect’ combinations.

attributive. See adjective.

auxiliary verb. In They were talking, the verb were is called an ‘auxiliary’ verb. This traditional term applies to a very small number of verbs such as be that can license ordinary verbs such as talk., that indicate grammatical categories such as tense or voice, or express meanings such as possibility and necessity, and that also have special grammatical characteristics which separate them from other verbs (called ‘lexical verbs’). These distinctive grammatical characteristics vary from language to language, and we list the English ones below.

English has three groups of auxiliaries:

1. be and have when used to form the progressive (She was wondering), perfect (I have paid) and passives (We were misinformed).

2. do when combined with another verb in an interrogative (Do you like it?) or negative (I don’t like it), or when contrasted with the negative (I do like it). Auxiliary do is used only when no other auxiliary verb is available, so it generally does not combine with other auxiliaries.

3. modal auxiliaries, which add meanings such as obligation and possibility (You must leave now. She may phone) or time (She will phone).

There are two main grammatical differences between these auxiliary verbs and all other verbs (including those like get and keep, as in We got talking or We kept talking). Both of these differences have something to do with polarity, the contrast between positive and negative.

  • negation: They can be combined with not or n’t, as in They weren’t talking (but not: *They gotn’t talking or: *She likesn’t to swim)
  • Closed interrogatives: They allow subject-auxiliary inversion, so they can be placed before their subject, as in Can she swim? (but not: *Got they talking? or: Likes she to swim?)

These distinctive characteristics provide an easy and clear distinction between auxiliary verbs and lexical verbs such as get that similarly license another verb as complement.

As explained in relation to clauses, grammarians disagree about the structural relations between an auxiliary verb and the next verb. They also disagree about the relation between lexical verbs and main verbs. For some, these two terms are interchangeable, but others distinguish them, so that a main verb may be a (non-lexical) auxiliary, as in She is tall or Have you a moment?

backshift. When It’s Tuesday today is reported as You said it was Tuesday today, the change of tense from present to past is called ‘backshift’. Notice how the past was no longer has its usual deictic meaning, because It was Tuesday today doesn’t make sense. Similarly, a direct past tense is backshifted to a past perfect: He arrived last week but She said he had arrived the week before. Backshift is found in some subordinate clauses that are subordinate to a matrix clause whose verb is in the past tense.

base. A word’s base is the morpheme, or combination of morphemes, to which affixes may have been added. For instance, in friendly and friends, the morpheme friend is the base, to which the affixes –ly and –s have been added.; but the word friend consists only of a base. Compound words such as blackbird are made up of two bases. A word’s base is sometimes called its ‘stem’, and in schools, bases are often called ‘root words’.

blend. A blend is a word derived from the start of one word and the end of another:
pictionary = picture + dictionary
smog = smoke + fog

motel = motor + hotel
brunch = breakfast + lunch

borrow, borrowing. The speakers of one language may ‘borrow’ words from another. For instance, the word tsunami is a borrowing (or loan word) from Japanese, meaning that English speakers use the word as if it was an ordinary English word, even if they know that it was originally Japanese. See also etymology.

cardinal numeral. See numeral.

case. In some languages, nouns and their modifiers have different inflectional forms, called ‘cases’, which reflect their grammatical function as subject, object and so on. English used to have a full ‘case system’, and German still has one which distinguishes four cases illustrated here for the noun phrases meaning ‘the old man’ and the plural ‘the old men’:




der alte Mann

die alten Männer


den alten Mann

die alten Männer


dem alten Mann(e)

den alten Männern


des alten Mannes

der alten Männer

Compared with English, it is striking how tightly integrated the inflection markers of case are with those of number (and gender, not shown here), and how these markers are distributed throughout the noun phrase.

In a typical case language such as German, a noun or noun phrase’s case is determined by its grammatical function:

  • subjects are ‘nominative‘ (e.g. Der alte Mann schläft. ‘The old man is sleeping’)
  • direct objects are ‘accusative‘ (e.g. Ick kenne den alten Mann. ‘I know the old man’)
  • indirect objects are ‘dative‘ (e.g. Ich gab dem alten Mann ein Geschenk. ‘I gave the old man a present.’)
  • modifiers of a noun are ‘genitive‘ (e.g. Ich kenne den Sohn des alten Mannes. ‘I know the son of the old man.’)

In other words, the rules of German require every subject to be nominative, every direct object to be accusative, and so on.

English is clearly very different from German with respect to case – most obviously, but certainly not only, in that the nominative and accusative forms apply only to a handful of pronouns and that there is no dative case at all. Because the differences are so great, some grammarians take the view that the category of case does not apply to English at all. Nevertheless, the traditional terms ‘nominative’, ‘accusative’ and ‘genitive’ are widely used in grammars of English, where the personal pronouns and WHO have four inflectional forms:

nominative I he she it we you they who
accusative me him her it us you them who/whom
genitive: dependent my his her its our your their whose
independent mine his hers its ours yours theirs whose

. cataphora. See anaphora.

class. See function, word class.

clause. A clause is typically a phrase headed by a verb, such as She came in or The dog ate my homework. Clauses are particularly important in grammar because typical examples (such as these) are potential sentences — i.e. they are grammatically complete. These are called ‘main clauses’ in contrast with ‘subordinate clauses’ which are grammatically subordinate and are often marked as such, as in Mary read a book while Jane wrote a letter, where while marks while Jane wrote a letter as subordinate. Another marker of subordination is finiteness, because only finite clauses can be main clauses, so to make any other kind of clause into a main clause it must be made finite. For instance, the first clause in Having no money left they slept in the park would need to be changed to a finite clause such as They had no money left.

The notion of ‘clause’ is also central to grammar because some of the most important classifications apply to clauses:

  • As mentioned above, clauses may be classified as main clauses or subordinate clauses and as finite or not.
  • They may be classified in terms of clause types such as ‘declarative’ and ‘interrogative’ (e.g. Mary read a book versus Did Mary read a book?).

Moreover, some of the most important syntactic functions are functions within clause structure. (though, depending on the preferred analysis, some of these functions may be found within clause-parts rather than clauses themselves).

The word-order rules that apply to these functions are one of the most salient characteristics of any language; for instance, English is commonly described as an ‘SVO’ language (where S and O stand for ‘subject’ and ‘object’, and V for the clause’s verb or verbs), meaning that the normal position of the subject is before the verb(s) whereas other parts of the clause, including objects, normally follow.

However, in spite of its central role in grammar, the notion ‘clause’ is problematic in relation to other grammatical concepts:

  • Phrase: In an example like Mary read a book, a clause can technically also be called a phrase whose head is a verb, so (matching the terminology of ‘noun phrase’, ‘preposition phrase’ and so on) we might expect to call it a ‘verb phrase’. However, this glossary avoids using this term because:
    • the term clause is too ancient and well established,
    • verb phrase has too many different meanings (illustrated below).
    • a clause may consist of nothing but the verb (as in Stop! or It stopped raining) so it does not technically count as a phrase.
    • in traditional school grammar, clauses had to be finite, while expressions like Mary borrowing my car were merely ‘phrases’ — a distinction rejected by most modern grammarians.
  • Sentence: In a single-clause sentence (such as Mary read a book. or What time is it?) the clause and the sentence are simply different descriptions of the same thing (much as a hut might be described either as a room or as a building).
  • Minor sentence. If a clause is a potential sentence, then minor sentences must also be ‘minor clauses’; but minor clauses break the usual rules for main clauses (e.g. Not to worry. Not that I mind.) and may even have no verb at all (e.g. How about a drink?).

clause structure. The structure of a clause is also a matter of major disagreements among grammarians. The main issue is whether there are intermediate structures between the clause and its verb or verbs. Consider the example Mary was reading a book, containing the auxiliary verb was and the lexical verb reading here are some alternative analyses that are each supported by a significant number of grammarians:

  1. Mary + [was reading a book]: subject + predicate (sometimes called ‘verb phrase’)
  2. Mary + [was reading] + a book: subject + verbal group (or ‘verb phrase’) + object
  3. Mary + was + [reading a book]: subject + verb + subordinate clause (or ‘verb phrase’).

This glossary assumes analysis 2 in some contexts, and 3 in others:

  • analysis 2 when discussing structure, e.g. the typical order of elements (SVO) in a clause.
  • analysis 3 when discussing licensing, e.g. the licensing relation between a verb and its complements (where reading licenses a book, but was is irrelevant).

(Compare the similar position regarding functions.)

clause type. Clause type is the technical term used for the following grammatical classification of clauses:

  • declarative: You are being tactful.
  • interrogative: further subdivided into:
    • closed interrogative:
      • yes/no interrogative: Are you being tactful?
      • alternative interrogative: Are you being tactful or just stupid?
    • open interrogative (or wh-interrogative): Why are you being tactful? Who came?
  • imperative: Be tactful. Let’s be tactful!
  • exclamative: How tactful you are being!

(A particularly important kind of interrogative in speech is the interrogative tag, as in This is yours, isn’t it?)

All but imperatives can also occur, usually with some difference in form, as subordinate clauses:

  • declarative: She told me that you are tactful.
  • interrogative: She’s not sure why you are tactful.
  • exclamative: She remarked how tactful you are.

We use each clause type for one characteristic speech act.

  • declarative: for making a statement
  • interrogative: for asking a question
  • imperative: for issuing a directive (request, order, instruction, etc.)
  • exclamative: for making an exclamatory statement.

But we also use them for many other speech acts such as invitations, suggestions, promises and threats. Similarly, the subtypes of interrogative clause typically ask for different kinds of answer (multiple choice for ‘closed’ and free choice for ‘open’), but other possibilities exist (e.g. which of the twins came? limits a potentially free choice to just two possibilities).

Grammatically, the types are distinguished by their initial structure:

  1. whether or not they have a subject. In imperatives, the subject is normally omitted but understood as ‘you’ (Come here!) but may be overt (You come here!). Another kind of imperative uses Let’s, with the subject understood as ‘we’ (Let’s go!).
  2. word order: normal or subject-auxiliary inversion
  3. the inflection of the verb
  4. the presence of a special word at the beginning of the clause.

These differences are summarised in the tables, starting with main clauses:

Imperative Declarative Interrogative Exclamative
1 optional subject subject subject subject
2 normal order subject-auxiliary inversion (or interrogative pronoun as subject) normal order
3 base form tensed tensed tensed
4 nothing nothing nothing or interrogative word what or how

For subordinate clauses, the structures are mostly the same, so the next table shows only the points of difference:

Declarative Interrogative
2 normal order
4 nothing or that If/whether or interrogative word

Because a sentence often consists of a single clause, the grammatical categories are often called ‘sentences types’, but clause type is the appropriate term because a sentence can consist of a combination of clauses of different types:

That book is mine, but whose is this one? (declarative + interrogative)
It’s cold outside, so do come in! (declarative + imperative)

I’m your friend, aren’t I? (declarative + interrogative tag)

cleft clause. An example such as It was Mary that I visited is called a ‘cleft clause’ because it is the result of ‘cleaving’ (splitting) a basic clause (I visited Mary) into two parts and then rejoining these parts with the help of the verb be. One part (Mary) is foregrounded while the other part is backgrounded as a relative clause (that I bought). There are two kinds of cleft clause:

  • In an ‘it-cleft’ such as the first example, the subject is the pronoun it and the backgrounded relative clause is put at the end: It + was + this book + that I bought.
  • In a ‘wh-cleft’ (or ‘pseudo-cleft’) the backgrounded part is a free relative clause, usually introduced by what, which is an ordinary noun phrase; so be simply links this phrase with the foregrounded element in either order: What I bought + was + this book or: This book + was + what I bought.

Cleft clauses are sometimes called ‘cleft sentences’, but they need not be complete sentences because they can be combined with other clauses: It was Mary that I visited, but I didn’t stay long.

clipping. One kind of abbreviation is produced by ‘clipping’ off one end of a word such as examination (to produce exam) or telephone (to produce phone).

cognate. See etymology.

coherent, coherence. An effective text needs to be ‘coherent’ and ‘cohesive’. The term coherence refers to the underlying logic and consistency of a text. The ideas expressed should be relevant to one another so that the reader can follow the meaning. The term cohesion refers to the grammatical features in a text which enable the parts to fit together (see cohesive devices). One way of creating cohesion is the use of anaphora: Emily sat down and turned on the television. Just then, she heard a strange noise. The phrase just then relates these events in time, and she relates Emily and the referent of she.

cohesive, cohesion. See coherent

cohesive device. Cohesive devices are words or other expressions that contribute to coherence. It could be argued that every word makes some contribution — e.g. the use of contribution in the previous sentence links back to contribute in the first sentence — but some words are strongly dedicated to this role. Such words show one or more of the following characteristics:

  • anaphora, e.g. So she did it.
  • definiteness, e.g. The girl liked the boy.
  • repetition, e.g. It’s his only wish, his only ambition, the only plan he pursues
  • lexical cohesion, e.g. She won easily. Victory was certain after the first five minutes.

Most word classes include some words which are cohesive, but some determiners and pronouns are especially important for building cohesion because they link anaphorically to earlier words. Some coordinators, subordinators and adverbs also make relations clear, as do some phrases Such words and phrases can be classified roughly by their meaning:
• addition: and, also, furthermore, moreover

• alternation: or, alternatively
• opposition: but, however, nevertheless, on the other hand
• reinforcing: besides, anyway, after all
• explaining: for example, in other words, that is to say
• listing: first(ly), first of all, finally
• result: therefore, consequently, as a result
• time: just then, meanwhile, later

Another kind of cohesive device is ellipsis, as in Fred’s story is more interesting than Bill’s, where Bill’s means Bill’s story.

collective noun. A collective noun such as team or crowd names a group of individuals which, in some sense, constitutes a single unit. English (especially British English), unlike some other languages, allows a singular collective noun to be treated as either singular or plural for purposes of agreement, depending on the sense. For example, The other team is more famous than ours. contrasts with: The other team are all taller than us. And plural agreement is the only possibility for a few collective nouns such as staff and police (e.g. The police are .. but not *The police is …).

collocation. A collocation is a pair (or larger group) of words that tend strongly to occur together, such as white wine, black coffee, dead easy.

comma. See punctuation.

comma splice. A comma splice is a pair of syntactically unrelated sentences punctuated, by a linking comma, as though they were a single sentence. For example: We caught our breath, something terrible had happened. This would be better punctuated as two separate sentences: We caught our breath. Something terrible had happened. Alternatively, the comma could be replaced by a colon (We caught our breath: something terrible had happened) or the wording could be changed to allow the comma (We caught our breath, for something terrible had happened.)

command. See speech act.

common noun. A common noun is a noun such as boy, book or beauty that can combine freely with:

  • a specifier such as the, a, some or my
  • a modifying adjective as in little boy or great beauty.

Common nouns are also either singular or plural (e.g. boy/boys, mouse/mice), and either count or non-count (e.g. thing, stuff). Contrast: proper noun and pronoun.

comparative. See grade.

comparative clause. See subordinate clause.

complement. Any constituent which a lexeme licenses is called its ‘complement’ (an ancient term alluding to the idea that a word’s grammar is not ‘complete’ without its complements). The term complement is the name of a grammatical function, contrasting with ‘subject‘ and ‘modifier’. A word’s complements are therefore the dependents that it ‘takes’ or ‘can take’, as when a transitive verb is said to ‘take’ an object or the verb DEPEND ‘takes’ on, whether these dependents are obligatory or optional. In contrast, a dependent which can occur optionally with any word with a suitable meaning and word class is not a complement but a modifier; so in We ate a pizza in the kitchen, a pizza is a complement of ate while in the kitchen is a modifier.

Traditional grammar has no general term for complements in this sense, but it does recognise three particular kinds of complement:

  • object (direct or indirect, e.g. She kept pigeons, She gave him a present
  • subject complement (e.g. She kept quiet)
  • object complement. (e.g. She kept the milk cool, where the milk is a direct object)

Modern grammars recognise other kinds exemplified by the underlined constituents in the following:

  • I’m waiting for some peace and quiet
  • I objected that there had been insufficient discussion
  • I hope to meet them in Bonn
  • I began feeling giddy
  • I helped wash up.

Such cases can generally be analysed satisfactorily simply as ‘complements’.

It is not only verbs that can have complements. Even in traditional grammar, prepositions also have ‘objects’ (e.g. in London) and in modern grammars complements are found with adjectives (e.g. fond of ice-cream) and with nouns (e.g. his reliance on excuses).

Many verbs license more than one kind of complement or various combinations of complements. Here, for example, is just a sample of those licensed by the verb keep:

  • The milk will keep better in the fridge (no complement)
  • He keeps pigeons (direct object)
  • I’ve kept you a seat (indirect object + direct object)
  • I’ll keep quiet (subject complement)
  • You must keep it warm (direct object + object complement)
  • I keep losing my keys (present-participial clause)
  • He kept on interrupting me (on + present-participial clause)
  • I kept them waiting (direct object + present-participial clause)
  • You are keeping me from my work (direct object + from phrase)
  • My income doesn’t keep up with inflation (up + with phrase)

The distinction between complements and modifiers is important for understanding punctuation, because a word is normally not separated by punctuation from its complements but may more easily be separated from its modifiers:

  • We enjoyed yesterday afternoon. [not: *We enjoyed, yesterday afternoon] — yesterday afternoon is a complement (more precisely, the object of enjoyed)
  • We met yesterday afternoon. [or: We met, yesterday afternoon] — yesterday afternoon is a modifier.

complex sentence. School grammar sometimes distinguishes three or four kinds of sentence: simple, complex, compound and compound-complex:
• A simple sentence consists of a single clause (e.g. It was raining.)
• A compound sentence consists of two or more coordinated clauses (e.g. It was raining but the sun was shining.)
• A complex sentence consists of a main clause with one or more subordinate clauses (e.g. If it rains, we’ll get wet.)

A compound-complex sentence consists of a combination of coordinated and subordinated clauses (e.g. If we don’t take an umbrella and it rains, we’ll get wet.)
However, this classification is misleading because ‘simple’ sentences can be complicated and ‘complex’ sentences can be very straightforward (e.g. I think you’re wrong.) A simpler classification contrasting ‘single-clause sentences’ with ‘multi-clause sentences’ may sometimes be helpful (provided that the uncertainties about clause structure can be resolved).

compound sentence. See complex sentence.compound tense. See tense.

compound word, compounding. A compound (or compound word) is a word made up of two base morphemes, e.g. football, headrest, broomstick, blow-dry, bone-dry. Compounding is a branch of lexical morphology, and relates lexemes in much the same way as derivation, except that the latter only involves a single base morpheme.

concrete noun. See abstract noun.

conditional clause. Conditional clauses are subordinate clauses typically introduced by the subordinator if and normally functioning as adverbial in the matrix clause, as in I will help you if you pay me. The subordinate clause expresses a condition under which the rest of the matrix clause is true: it gives a condition under which I will help you. Other subordinators used in conditionals are unless, providing, provided, in case and as long as.

There are two main types of conditional construction, ‘open’ and ‘remote’ (also called ‘real’ and ‘unreal’). Both can be used for conditions relating to future, present or past time:




If it rains tomorrow we will postpone the match.

If it rained tomorrow we would postpone the match.


If he is still here now, he will be in his office.

If he were/was still here now, he would be in his office.


If he bought it for that price he got a bargain.

If he had bought it for that price he would have got a bargain.

  • The open construction presents the fulfilment of the condition as quite possible, or even certain. The remote construction presents the fulfilment of the condition as a relatively remote possibility, or even impossible. Grammatically, open conditionals are the default type, with remote conditionals marked by two distinctive properties:The subordinate clause contains a past tense form or subjunctive were.
  • The matrix clause always contains a modal auxiliary – usually would, but should, could, might and, for some speakers, may are also possible.

Three of these six possibilities are often named and taught to learners of English as a foreign language as the ‘first’ (open future), ‘second’ (remote future) and ‘third’ (remote past) conditionals.

Some languages provide a special ‘conditional’ form of the verb meaning ‘would …’ (e.g. French chanterait, ‘would sing’), but it is unhelpful to extend this usage to the English combination would + verb.

conjunction. E.g. and, or, although, if. ‘Conjunction’ is one of the traditional parts of speech covering two subclasses, coordinating conjunctions and subordinating conjunctions. But modern grammars commonly treat these as distinct primary word classes (for which the obvious names are ‘coordinators‘ and ‘subordinators’), because they follow such different rules.

connective. ‘Connective’ is an informal name once used in schools for cohesive devices which serve to connect the ideas expressed in different clauses. The term cohesive device is preferable because it does not suggest a unified word class such as ‘noun’.

connotation. See denotation.

constituent, constituent structure. A constituent is a word or sequence of words that has a function in a larger grammatical unit. For instance, in one of the competing analyses of clause structure, His sister married a much older man the has three ‘immediate’ constituents:

These constituents in turn have their own immediate constituents:

  • His sister has two constituents, the specifier his and the head sister;
  • a much older man has three constituents, the specifier a, the modifier much older, and the head man.

Finally, much older has two constituents, the modifier much and the head older.

Such an analysis of a sentence from the top down yields what is called a constituent structure, which can be represented diagrammatically in various ways. One diagram system uses brackets:

[[His sister] married [a [much older] man]]

Another uses ‘tree structures’ in which each node of the tree (where two branches meet) stands for one constituent. Many such diagramming systems exist; for instance, in one system heads have a vertical link whereas their dependents have slanting links, as in this diagram.

This gives the full set of constituents in the sentence: these are the units that have to be described in terms of syntactic functions and classes. By contrast, there is nothing to be said about the sequence sister married: this is not a constituent. The more specific term ‘immediate constituent’ is often used when we are concerned with the construction in which the constituent has a function: much older, for example, is one of the twelve constituents of the sentence, but one of only three immediate constituents of the noun phrase a much older man.

The constituents in our example are all words or phrases. A clause can be a constituent provided it is not co-extensive with the whole sentence: in The man who came to dinner stole the silver. for example, the underlined subordinate clause is a constituent (it is modifier within the subject noun phrase) but the main clause which forms the whole sentence is not a constituent, not part of some larger construction. In addition, coordinations are constituents (again excluding a coordination of clauses that forms the whole sentence). Thus in I have been talking to [the boss and her husband] the bracketed coordination is a constituent functioning as object within the underlined preposition phrase.

content clause. See subordinate clause.

continuous. See progressive.

contract, contraction. The form I’m is a contracted form, or contraction, of I am. This type of abbreviation by elision is common in combinations of subject + auxiliary verb, and of auxiliary verb + not (e.g. can’t, won’t, hasn’t). Contractions of this kind are typical of informal speech and writing, and are reflected in spelling by the use of apostrophes.

convert, conversion. Conversion is a kind of derivation in which the derived lexeme has exactly the same base form but belongs to a different word class. For example, the verb walk (as in They walk fast) has been ‘converted’ into the noun walk, as in They had a nice walk. Conversely, the noun weed has been converted into a verb (We weeded the garden.). Conversion is still a particularly important process in English for creating new vocabulary (e.g. I’ll text you. We’ll action it later. That was very fun. Where’s the download?).

coordinate, coordinator, coordination. If words or phrases are coordinated, they are in a grammatically equal relation which is typically signalled by means of a coordinator such as and, or, but. (Contrast subordination.) The forms linked in this way may be of any length, from single words to complete clauses:
I bought the apples and bananas.

I bought some apples but they weren’t ripe.
In these examples the underlining shows the units that are coordinated, sometimes called ‘coordinates’; the sequence containing all the coordinates is called ‘a coordination‘. The coordinates need not be complete phrases; for example, they often consist of a whole clause minus its first few parts:

I have bought some apples and taken them home.

I went to a concert on Saturday and a play on Sunday. Another source of freedom is that a coordinate may itself contain coordinated elements:

I bought some apples and took them home but I forgot the potatoes. Coordination is a matter of grammar, not meaning, because the linked elements need not have equal status in meaning. For example, in the coordinated pair She worked all night and felt terrible next day, the feeling is a consequence of the working, so the meanings are not equal.

The grammatical equality of coordination can be seen in two ways:

  • The coordinates can share some element outside both, as in She speaks French and writes Spanish, where speaks French and writes Spanish have the same subject she, so that the example is equivalent to She speaks French and she writes Spanish. This is not possible in subordination; for example, She speaks French whereas she writes Spanish does not allow she to be shared, so we cannot say: *She speaks French whereas _ writes Spanish.
  • The coordinator belongs to the entire coordination rather than to one of the coordinates, so its position is fixed in relation to the whole structure: either between every pair of coordinates (e.g. apples and oranges and pears) or just before the last one, as in: apples, oranges and pears). It may be paired with an earlier ‘correlative’ coordinator just before the first coordinate (e.g. She both speaks Spanish and writes it). In contrast, a subordinator clearly belongs to the subordinate clause, so it moves around with this. For instance, starting with We got wet because it rained we can move the subordinate clause, including because, by fronting it to give Because it rained, we got wet. This is impossible with coordination (as shown by the ungrammaticality of *And we got wet, it rained, based on It rained and we got wet.)
    By these criteria, the coordinators of English are:
  • (both …) and: She both speaks French and writes Spanish.
  • (either …) or: She either speaks French or writes Spanish.
  • neither … nor: She neither speaks French nor writes Spanish. (contrast the adverb nor: …and nor does he …)
  • but: She speaks French but writes Spanish. (contrast the preposition but: She drinks nothing but tea)

These are the most important coordinators in English, but there are a few others which are either much less common, or less clear examples because they can be combined with and:

  • then: She first spoke French then wrote Spanish. (or: … and then …)
  • yet: She speaks French yet writes Spanish. (or: … and yet …)

Occasionally, the coordinator is omitted: He was [tired, hungry, bad-tempered].

Coordinations can occur at almost any place in the structure of sentences. In noun-phrase structure, for example, they can occur as:

  • specifier: [some or all] copies; [Sam’s and my] suggestions
  • pre-head modifier: an [original but somewhat implausible] idea
  • head: my [mother and father]
  • post-head modifier: students [who are very bright or who have wealthy parents]
  • complements: the revelation [that he was a pedophile and that he had only just been released from prison].

copula. When the verb be has a subject complement (e.g. She is clever, Jane is a linguist, it is commonly called ‘the copula’. See also linking verb.

corpus. A structured collection of extended examples of written or spoken English is called a ‘corpus’ (Latin ‘body’; plural: corpora). For example, the British National Corpus includes 100 million words of recorded spoken and written texts. Most corpora are stored on computer.

correct, correctness. See descriptive grammar.

correlative coordinator. See coordination.

count noun. Words such as pebble and patient are count nouns, in contrast with non-count nouns (sometimes called ‘mass nouns’) such as gravel and patience. As the terms suggest, count nouns can combine with the numerals one, two, three, etc., whereas non-count nouns cannot: one pebble, two pebbles, one patient, fifteen patients but not *one gravel, *two gravels, *one patience, *fifteen patiences. Count nouns thus always have an inflectional contrast between singular and plural, whereas non-count nouns do not. The great majority of non-count nouns have only a singular form, but there are some that have only a plural form (e.g. scissors, dregs, remains). A further difference is that count, but not non-count, common nouns have to be combined with a specifier.

We signal this contrast every time we choose some of the most common indefinite determiners such as a or some; for example, we say a pebble or a coin, but some grit and some money. Similarly, we use another, each, every, either, neither with count nouns, and much, enough, sufficient with non-count.

This contrast between individuals and material, or between ‘things’ and ‘stuff’, is an important grammatical resource as many noun lexemes can be used in either way, with correspondingly different interpretations; for example, we can distinguish a substance such as chocolate from particular portions or types of it, as in some chocolate compared with a chocolate.

dangling participle. See understood subject

dative. See case.

declarative. See clause type.

defining. See relative clause.

definite. Definiteness primarily concerns the referent of a noun or noun phrase though it can be extended as explained below. For example, the book is definite while a book is indefinite because they give the addressee different information about the book in question. The definite the book indicates that the book is assumed to be uniquely identifiable by the addressee, whereas a book indicates that it is not, so we use a book when referring to a book for the first time, and the book in subsequent mentions, as in I bought a book and a newspaper, and read the book on the train. Uniqueness reflects not only our knowledge of the preceding context (as in the previous example), but also our general knowledge; for example, we would use definite the father of my wife on the assumption that people have just one father, but indefinite a friend of my wife because they normally have more than one friend; and we would use the boss when the addressee knows which individual we are referring to (e.g. So what did you tell the boss?) but not otherwise (e.g. Sally’s got a new boss). In narrative, it is normal to use an indefinite when first introducing a character, replaced in later mentions by a definite: Once upon a time there was a king who had a horse. The king was very proud of the horse. … Definiteness applies in the same way to plurals: We met some nice people on holiday. …. The people we met on holiday were from Scotland.

As far as grammatical structures are concerned, indefinite is the default, so any noun or noun phrase is indefinite unless it is marked as definite in one of the following ways:

  • proper nouns without a specifier (e.g. I met John assumes a contextually unique and identifiable ‘John’; contrast He’s a real Rembrandt)
  • some specifiers:
    • the ‘definite article determiner the
    • demonstrative determiners: this/these, that/those)
    • genitives (e.g. my, John’s)
  • some pronouns:
    • personal pronouns
    • demonstrative pronouns

definite article. See article

degree. See grade.

degree modifier. Adjectives and adverbs that can vary for grade (also known as ‘degree’), may be modified by adverbs such as very, too, so and rather, or by nouns or noun phrases such as a bit or two inches, which are therefore called ‘degree modifiers’. Most potential degree modifiers of adjectives and adverbs can also function as adverbials, modifying a verb:

  • rather big/quickly/enjoy
  • a bit bigger/sooner, enjoy it a bit.

deixis, deictic. Words such as me, you, here and now have a special kind of meaning called ‘deictic’ (derived from a Greek word meaning ‘pointing’), which is based on the immediate context consisting of the speaker or writer (me), the addressee (you), and the time and place (here, now). This means that we cannot understand who or what such words are referring to (e.g. in I want to speak to you here and now.) unless we know the context in which they are uttered or written – who is speaking, who they are speaking to, and when and where they are speaking. Similarly, the tense of a verb is typically deictic since it locates a situation in relation to the time of speaking or writing. See also pronoun.

demonstrative. The determiners and pronouns this/these and that/those, which contrast ‘relatively near’ and ‘relatively far’ meanings, are called ‘demonstrative’. The distance involved may be purely spatial, or it may be more abstract, such as position in a text, e.g. this idea referring to an idea suggested in the immediately preceding text.

denotation. The denotation of a word is its basic meaning as given in a dictionary; for example, the denotation of book is the general class of things that we call ‘books’ (defined in terms of paper, words or pictures, reading and writing, pages and so on). In contrast, the word’s ‘connotation’ includes any additional values and associations the thing concerned may have for speakers, such as being exciting or boring; and its referent is the particular book referred to on a particular occasion.

dependent, dependency. See subordination.

derive, derivation. Derivation deals with one of the ways in which morphology can change words, and contrasts with both compounding and inflection. For instance, from friend we can derive friend-ly; from friendly, un-friendly; and from unfriendly, unfriendli-ness. Each of these examples is a different lexeme. In contrast, inflection changes the shape of a word to mark a grammatical feature, as in friend — friend-s (i.e. singular versus plural), but without changing it into a different lexeme. In English, words are derived by affixation (friend-ly), or conversion (verb walk → noun walk, as in have a walk), but we also use other more marginal changes such as blending (smog), abbreviation (Co., app) and acronym formation (AIDS).

It is better to reserve the term derivation for relations between words within one language and at one time, in contrast with etymology, which relates a word to its historical origins in the same language or a different one.

descriptive. Grammar and vocabulary may be studied in two ways:

  •     ‘Descriptive’ analysis describes (and tries to understand) grammar or vocabulary as it actually is, in the usage of native speakers. Descriptive methods apply equally to Standard and non-Standard English, and indeed to any language, whether or not it has a ‘standard’ form; and because different varieties have different rules, descriptive analysis recognises and describes variation. For instance, for a descriptive grammar, ain’t is not allowed in Standard English but is allowed in most non-Standard varieties. Where usage is divided or uncertain, descriptive analysis records the alternatives without making judgement.
    Most comprehensive grammars and dictionaries are descriptive, and are based on careful study of the actual usage of expert native speakers and writers. These descriptions provide guidance, mediated by textbooks and style guides, for non-experts such as second-language learners or native speakers still developing mature competence as writers or speakers. Language description can even underpin small-scale language reform aimed at improved communication such as the present glossary of recommended grammatical terminology.
    In contrast, ‘Prescriptive’ analysis does not describe usage, but tries to change it by ‘prescribing’ or recommending some forms as absolutely ‘correct’ and ‘proscribing’ others as absolutely ‘incorrect’ or ‘wrong’, or as ‘errors’. In principle, this could include harmless language reform, but in practice, the term as used by linguists is pejorative. For most linguists, a rule is prescriptive if it rejects a form as incorrect under all circumstances even when it is actually widely used; so the main objection to such rules is that they ignore variation. Proscriptions are typically based on spurious arguments such as logic (Don’t use negative concord), the grammar of other languages (Don’t split infinitives), unnecessary avoidance of ambiguities (Don’t use dangling participles) or standardness (Don’t use ain’t). At one time, prescription dominated the teaching of grammar, where it often resulted in a list of ‘common errors’ which children learned to avoid; but (partly under the influence of linguistics) this is much less so now.

determiner. E.g. the, a, this, any, my. A determiner stands before a noun and any words that modify the noun, and has the function ‘specifier’ in the noun phrase. Determiners are said to ‘determine’ the meaning of the noun, rather than to ‘modify’ it, because they carry special types of meaning such as definiteness, number and countability.
Determiners include:

  • articles: a/an, the
  • demonstratives this/that, these/those
  • quantifiers some, any, no, every, each, either, neither; and (for some grammarians) many, much, few, little, enough, all, both
  • some interrogative wordswhich (which car?), what (what size?), whose (whose coat?)
  • the relative whose (e.g. the person whose name was on the door) and which (by which time)
  • the singular numeral one
    Some grammarians also classify as determiners plural numerals although they never combine with singular count nouns like boy, and some can combine freely with other determiners (e.g. his three cats, these many years).

Most words used as determiners can also be pronouns. These include:

  • the demonstratives: I bought this.
  • question words: Which did you buy?
  • most of the quantifiers: I bought some.
    When they are pronouns, these words are not followed by a noun, although their meaning may include an understood noun which is provided by the context:
    I’ve got some books. This (i.e. this book) is for you.

dialect. See variety.

direct object. See object.

direct speech. There are two ways of reporting what somebody says: direct speech and indirect speech.

In direct speech, we use the speaker’s original words (as in a speech bubble). In a written text, speech marks (‘…’ or “…” – also called ‘inverted commas’ or ‘quotes’) mark the beginning and end of direct speech:

Helen said, ‘I’m going home now’.

What do you want?” I asked.

“Will it ever end?” I wondered.

Direct speech is analysed as a text, rather than as a subordinate clause (e.g. as subject or object); so it may consist of any number of sentences, or even an incoherent jumble of words in different languages.

  • The agent said, “Alors — welcome, monsieur. Bravo. Well well.”

In indirect (or reported) speech, we report what was said by integrating it into our own sentence structure, without necessarily using the exact words of the original speaker. Speech marks are not used, and deictic elements such as pronouns, time and place references and verb tenses may be different from those used by the original writer or speaker to reflect the changed standpoint:

Original utterance/thought

Direct speech

Indirect report

I am going home tomorrow

Helen said, “I am going home tomorrow”

. Helen said (that) she was going home the next day.

What do you want?

I said, “What do you want?”.

I asked them what they wanted.

Will it ever end?

” Will it ever end?”, I wondered.

I wondered whether it would ever end.

directive. See speech act.

discourse. Any continuous stretch of language in ordinary use, whether spoken or written, can be called ‘discourse’ – a mass noun. A particular example of discourse is called ‘a text’ (a count noun). A text may involve a single person as speaker or writer, but it may also involve a group of interacting people. Discourse is controlled by rules and conventions that go beyond the rules of grammar, such as the need for coherence and cohesion.

double negative. See negative concord.

dummy. See pronoun.

en form. See participle.

elide, elision. If a sound is omitted in informal speech it is said to be ‘elided’; e.g. t can be elided in want to to give wanna. Eliding sounds produces contracted words, as when and contracts to’n in boys ’n girls. Contrast ellipsis.

ellipsis, elliptical. The process called ‘ellipsis’ allows the omission of words in order to avoid repetition; for example, Mary doesn’t like oranges, but Jane does is understood as meaning ‘does like oranges’. This is a type of anaphora because it links the ellipsis to its antecedent (like oranges).

Ellipsis is possible with some members of most word classes:

  • verbs
    • all auxiliaries, e.g. Does anyone like swimming? Yes, Jane does, doesn’t she? (where doesn’t she? is an interrogative tag).
    • many other verbs whose complement is optional, e.g. I don’t agree. She has already left.
  • prepositions, e.g. She opened the box and looked inside.
  • common nouns, e.g. There were some nice apples so she bought a pound.
  • adjectives, e.g. I’m not interested. The resulting turmoil soon subsided.
    • including comparative and superlative inflectional forms, e.g. This painting is good, but that one is better, and this one is the best.
  • comparative and superlative adverbs, e.g. I’d rather have a beer. Mary drove faster.

etymology. A word’s etymology is its history: its origins in earlier forms of English or other languages, and how its form and meaning have changed. For example:

  • our word thatch is etymologically related to the German word Dach, which means ‘roof’ (whether thatched or not), so thatch and Dach are called ‘cognates’ because they are different forms of the same word in an earlier language from which both English and German are descended.
  • our word verb is related to the Latin word verbum, which means simply ‘word’ and also lies behind our adjective verbal; since we borrowed verb from Latin, it is called a ‘loan word’

euphemism. A mild or vague expression used instead of one with bad connotations; for example, pass away is a euphemism for die.

exclamation. An exclamation expresses strong emotion (joy, wonder, anger, surprise, etc) and, if written down, it is usually followed by an exclamation mark. Exclamations can be full sentences expressing ‘exclamatory statements’ (e.g. That’s a real blow!), phrases (e.g. That Smith woman!), formulaic expressions with specialised meanings (e.g. Oh dear! Good grief!) or specialised words called ‘interjections’ such as aha! and ow! Exclamations are sometimes expressed as exclamative clauses such as What a nice house you have!

exclamative. See clause type.

existential clause. An existential clause is one in which the subject is the pronoun there, as in There’s a fly in my soup, or (more formally): There arose a serious dispute. In these clauses there replaces a noun or noun phrase which could have been the subject, but which stands after the verb: a fly or a serious dispute; and they are called ‘existential’ because their main point is the existence of whatever is referred to by the delayed potential subject — the fly or the dispute.

exophora. See anaphora.

extrapose, extraposition. A subordinate clause may often be ‘extraposed’ (‘moved out’) from its expected position to the end of the larger clause; for example, the following examples mean exactly the same, but the subordinate clause (underlined) is replaced in the second example by the pronoun it, which allows the clause to be moved to the end of the sentence.

  • Whether my scheme would work is still unclear.
  • Ii is still unclear whether my scheme would work.

feminine. See gender.

filled pause. See word class.

filler. See word class.

finite. In some languages, a verb‘s inflectional forms fall into two groups according to whether or not they have a limited range of subjects; in Latin, for instance, amo, meaning ‘I love’, is limited to first-person subjects whereas amare, ‘to love’, isn’t. The traditional term finite (literally, ‘finished’ or bounded) evokes this limitation. Finite verbs are important because they are required by a main clause — in other words, any sentence must normally contain at least one finite verb. Conversely, non-finite verbs such as participles and infinitives are used to mark a clause as subordinate.

Finiteness is important for English as well, but applies to clauses rather than to verb-forms as such. So in English, finite clauses typically have to have a subject and can be used as the only verb in a main clause. These clauses have either present-tense or past-tense, so they are sometimes called ‘tensed‘. Finite clauses may also be imperative (found in main clauses but normally having no overt subject) or subjunctive (having an overt subject but not found in main clauses). In short:

  • finite
    • present tense: We take our time. She takes her time.
    • past tense: We took our time.
    • imperative: Take your time!
    • subjunctive: I recommend that she take her time.
  • non-finite
    • present participle: I found her taking her time.
    • past participle: (We have taken our time)
    • passive: The time taken to do it was half an hour.
    • infinitive: I want to take my time..

(The ‘past participle’ example is relevant only in some analyses of clause structure.)

formal. Formal style is a register which is typically reserved for occasions which are public and impersonal, in contrast to the informal style typical of everyday speech. For example, beginning a question with a preposition before the wh-word is a formal option in English: To whom does the coat belong? (Contrast the more informal style of Who does the coat belong to?).

free relatives. An ordinary relative clause is separate from the noun that it modifies; so in the thing which I made, the relative clause which I made is separate from the head noun thing, which is the antecedent of which. In contrast, a free relative (also called ‘fused relative’) combines the head noun and the relative pronoun into one; so what I made is not only semantically but also syntactically equivalent to the thing which I made (or, much more formally, that which I made) In this analysis, the words the thing which (or that which) are ‘fused’ into the word what, so the whole phrase is a noun phrase.

Free relatives are introduced by a special range of relative pronouns, notably what:

  • I paid for what I bought (what I bought = the thing that I bought)
  • What I bought was a book.

(The second example illustrates one of the main uses of free relatives, in a ‘wh-cleft sentence’.) However other pronouns and determiners can also introduce free relatives:

  • Whoever tied this knot tied it too tight.
  • What books he had were in the attic.

Free relatives are easily confused with subordinate interrogative clauses which can start with the same pronouns. For example, what he made could be either:

  • free relative: We ate what he made = We ate the food that he made.
  • subordinate interrogative: I wonder what he made = “What did he make?”, I wondered.

front, fronting. A constituent that would normally follow the verb may be ‘fronted’ to the start of the clause; for instance:

  • These models we can supply within three days. (fronted object; compare: We can supply these models within three days)
  • Last night, the wind blew our tree down (fronted adverbial; compare: The wind blew our tree down last night).
  • We were sent to learn, and learn we did.(fronted lexical verb; compare: We did learn).

See word order.

function. Grammatical analysis distinguishes classes (such as ‘noun’ or ‘noun phrase’) from functions such as ‘subject’ or ‘object’. Whereas a constituent‘s class relates it to the rest of the language, its function relates it to the rest of the sentence. For instance, in The girl caught a fish, we classify both the girl and a fish as noun phrases, which recognises, among other things, that they could both be used either as a subject or as an object; but in this particular sentence, the girl has the function ‘subject’ while a fish functions as ‘object’.

Here are some consequences of this fundamental distinction between classes and functions:

  • In terms of single words, a dictionary lists each lexeme’s primary word-class (as ‘noun’, ‘preposition’, etc), but cannot list its function because this varies from sentence to sentence.
  • As explained above, a single class may be available for several different functions; e.g. noun phrases may function as subject or object.
  • Conversely, a single function may be available for two or more classes; for instance, , in recent government inquiry, both recent and government have the same function (modifier of inquiry), but although recent is an adjective, government is a noun.

Functions can themselves be classified into very general categories which can in turn be broken down into more specific ones. A constituent’s function shows how it combines with other constituents to form a phrase, so the first split is between the phrase’s head and all the subordinate constituents. The next split distinguishes different dependents according to how the grammar allows them to combine with the head. This split produces four further general categories, so we have five general types of function:

One technical issue is whether a constituent’s function relates it to the containing phrase or just to its head. For example, in The dog buried a bone in the garden, the subject is the dog, but does this mean that it is the subject of the whole clause, or just of its head, the verb buried? Grammarians are divided on this question, and this glossary takes both approaches in different places.

  • In discussing licensing, the functions relate to the head because this is the licenser; so we can say that because bury licenses an object, a bone is the object of buried
  • In discussing structure, the functions belong to the whole containing phrase; so our example is a clause with the structure ‘subject — head — object — modifier’.

Grammatical functions are different from semantic roles, but are closely related to them because a constituent’s grammatical function is a guide to its semantic role. For instance, when a verb licenses an object, it generally determines the latter’s semantic role; so break licenses an object referring to the patient (the thing affected), while make licenses one which refers to the thing created.

function word. See grammatical word.

fused relative. See free relative.

future tense. Like many other languages, English has several ways of referring to future events. The principal ones are:

  • shall/will + base-form: I shall/will probably be away next week. Nobody will ever know the truth about his disappearance.
  • a presenttense: We leave on Tuesday
  • a present progressive: I am seeing Jake tomorrow.
  • be going + toinfinitival complement: They are going to dig up the road again.

Some languages have a future tense comparable with the present and past tenses. For example, French contrasts chantera, ‘will sing’, with chante, ‘sings’ and chanta, ‘sang’. Traditional grammar treats English as having a future tense too, with a three-term system contrasting future will sing, present sing/sings and past sang. Modern grammars, however, generally reject this analysis, treating will as a modal auxiliary like can, may, must, etc., rather than as a future tense auxiliary.There are two main arguments. In the first place, will is itself a present-tense verb with would as its past-tense counterpart: compare I’ve asked him to help, but he won’t and I asked him to help but he wouldn’t. This means that we have four terms to consider, not just three:


Other verb will
Present tense He is king He will be king
Past tense He was king He would soon be king

The lexeme will thus doesn’t contrast with present and past tense but combines with either.

Secondly, will belongs with the modal auxiliaries not only grammatically but also semantically. The meaning difference between will be and is is not future time vs present time, as evident from pairs like She will be free tomorrow and She is free tomorrow (both future time) vs She will be in Paris now and She is in Paris now (both present time). Moreover, the modals can, may, must can also be used for both future and present time: She may be free tomorrow, She may be in Paris now.

gender. When used as a purely grammatical term, gender refers to the contrast that some languages make between categories such as masculine, feminine and neuter; for instance, the masculine or feminine gender of a French noun determines whether the definite article has the form le or la, and adjectives agree in gender with their noun. When applied to humans and some animals, grammatical gender generally corresponds to the biological contrast of sex, but for inanimates it is basically arbitrary; so in German, ‘spoon’ is masculine while ‘fork’ is feminine and ‘knife’ is neuter, giving ein grosser Löffel (a big spoon) but eine grosse Gabel (a big fork) and ein grosses Messer (a big knife). In contrast, languages such as English have no such classification of nouns, but grammarians often describe the personal pronouns he/him, she/her and it as having masculine, feminine and neuter gender

generic reference. Nouns and noun phrases can be used to refer either to specific individuals (e.g. The dog barked) or (generically) to an entire class or species (e.g. The dinosaur is extinct). The latter use is often called ‘generic reference’

genitive nouns and noun phrases. In grammars for some other languages, genitive is the name of a case (e.g. in Latin, Romuli amicus , ‘friend of Romulus’, Romuli is the genitive form of Romulus; in German, das Haus des Mannes, ‘the house of the man’, where des and Mannes are both genitive); the genitive case is used, among other things, for the modifier of a noun.

  • Although it is less clear that English has case, the term genitive can be applied to nouns or noun phrases to which the apostrophe and -s are added (e.g. Mary’s house, the dog’s tail, the girl next door’s name) typically used as specifier inside a noun phrase. 

School grammars use the term ‘possessive’ rather than ‘genitive’, on the basis of examples like Mary’s watch or Mary’s stamp collection, where we understand that Mary possesses the watch or stamp collection. ‘Possessive’, however, is a misleading term because there are so many commonplace examples where there is no possessive meaning:

  • Mary’s father
  • Mary’s birth
  • Mary’s school
  • Mary’s anger
  • Mary’s lack of money
  • Mary’s refusal to compromise
  • Mary’s rejection of the offer
  • Mary’s rejection by the interviewer

Personal pronouns also have genitive forms. When used as specifier, they are: my, your, his, her, its, our, their; and when used without a following common noun they are: mine, yours, his, hers, its, ours, theirs. (A similar alternation is found in the determiner no, which alternates with none.)

When the -s of the genitive is added to a plural noun which already ends in plural -s, the two suffixes merge in one; so we find the children’s names, but the boys’ names. The spelling generally follows this rule of pronunciation.

A related construction, which allows many of these meanings, uses the preposition of, which is often preferable especially with inanimate or long objects. For example:

  • The fall broke Mary’s arm but: She broke the arm of the chair.
  • their arrival but: the arrival of the Queen of Sheba

See also group genitive.

gerund. When a clause headed by an ing-form verb is used like a noun phrase, it is a ‘gerund’, in contrast with similar clauses used as participles. For example:

  • gerund: Drinking alcohol at work is forbidden.
  • participle: Anyone drinking alcohol at work will be dismissed.

Gerunds provide an important resource for language users by allowing clausal meanings to be ‘nominalised’.

govern. See license.

gradable, grade. Gradable adjectives such as big and quick define qualities that people or things may have to different degrees. One of the grammatical characteristics of gradable adjectives is to allow comparisons, as in This is bigger than that or This is the biggest of all. The inflectional forms bigger and biggest are called ‘comparative’ and ‘superlative’, and the contrast between these and the ‘positive’ (or basic) form is called ‘grade’ or ‘degree‘. A similar contrast applies to adverbs such as soon — sooner — soonest. This inflectional contrast only applies to short adjectives and adverbs; longer words use more and most instead, as in more/most important/importantly.

grammar. The term grammar is used in two main senses: as the name for a part of language, and as the name for the study of this.

1. The grammar of a language consists of that language’s general rules for using words, and sentences built out of them, to express meanings. Like any other social conventions, it can be seen as a collection of restrictions (Don’t do X) or of permissions (Doing X is ok) – i.e. as limiting or as enabling. Many linguists contrast a language’s grammar with its vocabulary (or ‘lexicon‘) and with its phonology (sound system).

Grammar is normally divided into two parts:

2. Grammar is the study of grammars in sense 1 above. There are two different approaches to the study of grammar, called descriptive and prescriptive.

Related words are grammatical (as in a grammatical study — but not This sentence is grammatical, for which see the next entry) and grammarian (one who studies grammar). A grammar is, of course, a book describing the grammar of a language (sense 1), one possible output of the study of grammar (sense 2).

grammatical. A sentence is described as ‘grammatical’ if it is allowed by the grammar of the language or variety concerned, and as ‘ungrammatical’ if it is not. For example, *I him see is ungrammatical in all varieties of English, while I ain’t seen nobody is grammatical in some varieties of English but not in others.

Grammarians write ‘*’ before examples to show that they are ungrammatical, and ‘?’ when they are uncertain.

grammatical word. Some grammarians call words such as to, that, the and is ‘grammatical words’, in contrast with ‘lexical words’ such as dog or bark. Like the contrast between grammar and lexicon, this is a matter of degree, without any clear boundary, but it can be helpful in thinking about grammar.

greeting. See word class.

group genitive. If we add the genitive apostrophe ’s to a noun which has a following modifier, such as girl with brown hair, the result (e.g. the girl with brown hair’s name) is called a ‘group genitive’ because it is clear that the apostrophe belongs to the entire phrase (a ‘group’ of words) rather than just to the head noun (girl).

head. A phrase consists of one word, its head, and all the other constituents which are subordinate to this word. As such, the head is the syntactically most important word within the phrase, the one which determines what class of phrase it is (if the head is a noun, the phrase is a noun phrase, if the head is an adjective the phrase is an adjective phrase, and so on) and hence what functions it can have in larger constructions. Since all the other constituents of the phrase are subordinate to the head, they are irrelevant to its classification and function. For example, in the adjective phrase very silly, the head is the adjective silly, and in the noun phrase very silly squirrels the head is the noun squirrels.

Although the head word is grammatically important, as the word to which all the other words are subordinate, it need not be the most important in terms of meaning and information; for example, in She’s a fascinating person, the phrase a fascinating person has the rather predictable and unimportant word person as its grammatical head.

‘Head’ is a function label and can be applied to words of different word classes. Because the word class of the head decides how the whole phrase is classified, we can distinguish adjective phrases, noun phrases, preposition phrases and adverb phrases. But for phrases headed by verbs (or verbal groups), this glossary recommends the term clause.

historic present. One of the uses of the present tense is to describe an incident in the past as though it was happening now: I’m sitting in this cafe, and a guy comes up to me and says …

homograph. If two different words have the same spelling as each other, they are often called homographs:

The bear growled – I can’t bear it. (same pronunciation, so also homophones)

a lead pencil — the dog’s lead (different pronunciation)
The term homograph is sometimes reserved for words that have the same spelling but are not homophones, like lead above. See also homonym.

homonym. Words which are either homophones or homographs are often called ‘homonyms’. The term homonym is sometimes reserved for words which are both homophones and homographs (such as the noun and verb bear). It also tends to be reserved for words whose meanings are radically different, such as bank (edge of a river, or a financial institution), in contrast with polysemous words, whose alternative meanings are closely related (e.g. the verb bear is polysemous because it can mean either ‘carry’ or ‘tolerate’ and these meanings are related).

homophone. Two words are homophones if they have the same pronunciation (but, of course, different meanings):
The fair has arrived. She has fair hair. (same spelling)

read – reed (different spelling)
The term homophone is sometimes reserved for words that have the same pronunciation but are not homographs. See also homonym.

hypernym. See hyponym.

hyponym. The word dog is described as a hyponym of animal, and conversely, animal is a hypernym of dog because a dog is a subcategory (or type or kind) of animal. (Hint: in Greek, hypo means ‘beneath’, as in hypodermic, ‘under the skin’; hyper means ‘above’, as in hypermarket; Greek hypo and hyper correspond to Latin sub- and super-) The hypernym’s meaning includes the meaning of the hyponym, so the hypernym’s meaning is more general than the hyponym’s. Words are often linked in a chain of increasingly specific meanings; so although dog is a hyponym of animal, it is a hypernym of terrier.

idiom. An idiom is an expression which is not meant literally and whose meaning cannot be deduced from knowledge of the individual words. For example:
You look a bit under the weather this morning.
Are you
all right?
You’re always introducing
red herrings.
You and I have the same problems — we’re
in the same boat.
That name
rings a bell. I’ve heard it before somewhere.

I’m going to pack up now, so you’d better leave too.

imperative. See clause type. (See also command.)

imperfect. The term ‘imperfect’ is used in grammars of languages such as French for the verb form that might be translated by the English past progressive (e.g. was singing) or by used to (e.g. used to sing). The imperfect is not the opposite of the English perfect, so English has no verb form or construction that could be called ‘imperfect’.

indefinite article. See article.

indefinite determiner or pronoun. See definite.

indicative. See mood.

indirect object. See object.

indirect speech. See direct speech.

infinitive, infinitival. In I want to gohome, the clause (to) go home is called ‘infinitival’. In English, an infinitival clause has a verb with base-form morphology and is syntactically subordinate (e.g. I did it to please her). In languages such as French, a verb’s ‘infinitive’ is the name for an inflectional form (e.g. porter, ‘to carry’, with –er added to the base port). English infinitival clauses often have to before the verb, so this is sometimes treated as ‘part of the infinitive’, leading to the prescriptive ban on ‘split infinitives’ (as in to boldly go …); however, it is very clear that to is not part of the verb, though of course it is part of the clause containing it.

inflection, inflectional form. Inflection has to do with variation in the form of lexemes. The adjective lexeme TALL, for example, has the three forms tall, taller and tallest: these are called ‘inflectional forms’. Similarly, the noun lexeme CHILD has two inflectional forms: child and children. There are two sides to inflection, its affects on the word’s morphology and on its syntax:

  • Inflectional morphology deals with the actual differences between the forms: tall is identical with the base of the lexeme, and taller and tallest are formed by adding the suffixes -er and -est to the base.
  • In relation to syntax, the rules of syntax apply differently to different inflectional forms. For example, tall, taller and tallest follow the rules for positive, comparative and superlative adjectives, so we can say He’s taller than you, but not *He’s tall/tallest than you; He’s the tallest of them all, but not *He’s the tall/taller of them all. In contrast, lexical morphology does not create syntactically special classes; for instance, worker, formed by adding the suffix —er to the base, is syntactically just the same as any other common noun such as man.

Different inflectional forms must be formally distinct in at least some lexemes; for example, He tried to be patient and Be patient! must contain the same inflectional form of be, in spite of the syntactic and semantic differences between infinitivals and imperatives, because the inflectional forms are always the same, even in the highly irregular verb be.

Applying this principle to English verbs, modern grammarians generally recognise six inflectional forms, though different grammarians use different names for them. The differences can be seen most clearly in some irregular verbs such as take, and especially in the verb be, so they are illustrated below with be and take as well as with the regular walk:

  • was/were, took, walkedpast form, found in past-tense finite clauses.
  • is, takes, walks — s-form, found in present-tense finite clauses with third-person singular subjects.
  • am/are, take, walk — plain present form, found in present-tense finite clauses with subjects other than third-person singular.
  • be, take, walk base form, found in imperative, infinitival and subjunctive clauses, and after most modal auxiliary verbs.
  • being, taking, walking — ing-form, found in progressives, present-participle clauses and gerunds
  • been, taken, walkeden-form found in perfects and in passive clauses. The term ‘en-form’ reminds us that some irregular verbs such as take and see have a form with the suffix -en (taken, seen) which is distinct from their past tense (took, saw). In a regular verb, the en-form is always the same as the past form.

English grammar has considerably fewer inflections than languages such as French, German and Latin, but inflection is an important part of English grammar.

informal speech. An informal (or ‘colloquial’ or ‘casual’) register is language used in familiar, informal contexts, in contrast with more formal registers. For instance, How about a drink? is informal, in contrast with Would you care for something to drink? (See also: active verb, contraction, phrasal verb, preposition, tag question)

ing-form. See inflect.

instrument. Instrument is the name of a semantic role. In She broke it with a hammer, the hammer (not the word) is called the ‘instrument’, in contrast with the agent (her) and the patient (it).

intensifier. Words such as thoroughly and hardly, as in thoroughly disapprove or hardly know, are sometimes called ‘intensifiers’, but a better term is degree modifier.

interjection. See exclamation.

interrogative. See clause type.

interrogative tag. (Often called ‘tag questions’) One of the characteristics of speech is the use of small elliptical interrogative clauses at the end of a main clause: That boy can swim well, can’t he? You missed the deadline, didn’t you?

The tag has the form of an elliptical interrogative clause, usually consisting of just two words: an auxiliary verb and a personal pronoun subject. The auxiliary matches that of the main clause if it has one (He has gone home, hasn’t he?); otherwise it is the matching form of the dummy auxiliary do (She enjoyed it, didn’t she?). The pronoun is anaphoric to the subject of the main clause (Her son spoke well, didn’t he?; The students haven’t met her, have they?; I’m invited too, aren’t I?)

Interrogative tags affect the speech act performed by someone using the clause to which they are added. The polarity of the tag is usually the reverse of that of the main clause, giving positive + negative or negative + positive, as in the above examples. Such tags are typically used to encourage agreement (e.g. It’s getting colder, isn’t it?) or to seek confirmation (e.g. We don’t have to pay in advance, do we?).

It is also possible to keep the polarity constant: Your husband is coming too, is he? These tend to have some emotive overtone, of surprise, disapproval, disbelief or the like (So the dog ate your assignment, did it?)

Intonation may also affect the meaning, as in This mess is yours, isn’t it?, where the intonation on the tag may either rise or fall according to whether the speaker expects the addressee to agree.

Variable interrogative tags are a peculiarity of spoken English. In many languages the tag has a constant form (e.g. in French: n’est-ce pas; compare English innit).

interrogative word. (Sometimes called ‘question word’) These are determiners, pronouns and adverbs that are used to introduce interrogative clauses.

  • Which book did you read?
  • What happened?
  • How did you do it?
  • To whom should I address the letter?

Most interrogative words in English are spelt with wh… (who, what, which, when, why, where, whoever, etc.), so, along with the relative pronouns, they are sometimes called ‘wh-words’. (The exception is how, but even this contains w and h.)

intransitive verb. See transitive verb.

invert, inversion. See auxiliary verb.

irregular. See regular.

it-cleft. See cleft clause.

lexeme. See word.

lexical morphology. See morphology.

lexical verb. See auxiliary verb.

lexical word. See grammatical word.

lexicon. A language’s lexicon is its vocabulary – i.e. its stock of words (or more precisely lexemes – ‘dictionary words’). Traditionally this is contrasted with the language’s grammar, but there is no clear boundary between the fine detail of one and the broad generalisations of the other.

license. The head of a phrase ‘licenses’ (or ‘governs’) its complements. A complement is a dependent that is permitted with some but not all of the lexemes that function as head of the phrase. For example, an object can occur with the verb DESTROY but not with FAINT: we can say They devoured it, but not *They fainted it. So DESTROY is said to license an object, whereas FAINT does not. Modifiers, on the other hand, are not licensed. A modifier of time, for example, can occur with any verb, so the admissibility of yesterday in They devoured it yesterday is not due to some licensing property of DEVOUR, but simply to its being able to modify and verb .

A non-technical term equivalent to license is take: ‘DEVOUR takes an object, but FAINT doesn’t’.

Licensing determines one or more of the following properties of a complement:

  • whether it is optional or obligatory (e.g. an object is optional with eat but obligatory with devour)
  • its function (e.g. direct object)
  • its word or phrase class (e.g. discuss requires a noun or noun phrase as its object)
  • its lexeme (e.g. depend selects the preposition on)
  • its inflectional class (e.g. will selects abase form but have selects an en-form; and in some languages, the head selects the object’s case).
  • its semantic role (e.g. the object of break is the ‘patient’, whereas the object of make is the thing created)

Verbs and nouns also determine similar properties of their subjects and specifiers; for instance, rain requires its subject to be it and Sudan requires the specifier the. But in these cases, the dependents’ properties are also determined by the head’s inflectional form; for instance, whether the subject is required or merely possible depends on whether the verb’s form marks its clause as finite.

linking verb. Verbs that are used with a subject complement are sometimes called ‘linking verbs’ (or ‘link verbs’). The most common example is the copula be, but others are become, seem, look (as in They look nice) and get (as in You’ll get wet).

loan word. See borrowing.

main clause.

A clause is a main clause unless it is a subordinate clause.

.main verb. The last verb in a clause is normally its main verb:

  • The dog buried a bone.
  • The dog has been burying a bone.
  • The dog is a spaniel.
  • I haven’t a clue.

As these examples show, although the main verb is normally a lexical verb such as bury, it may be an auxiliary verb such as be or have. Exceptionally, the main verb may be missing through ellipsis, as in Yes, I will in answer to Will you come back soon?

masculine. See gender.

mass noun. See count noun.

matrix clause. A clause containing a subordinate clause is called the ‘matrix clause’ for that clause. For example, in I told you that I would come if I could,

  • the matrix clause for if I could is the subordinate clause that I would come if I could
  • the matrix clause for that I would come if I could is the main clause I told you that I would come if I could.

As can be seen from this example, a matrix clause need not be the main clause.

metalanguage, metalinguistic. Metalanguage is the language we use when talking about language itself. It includes words like sentence, noun, paragraph, meaning, pronunciation and all the terms in this glossary.

metaphor. A metaphor is an expression which uses the literal meaning of words to create a so-called figurative meaning that resembles the literal meaning in some way, for example describing anger in terms of fire or heat:
She was boiling with rage.
She is an ass.
She fell in love.

minor clause. A typical main clause is headed by a verb. In contrast, a ‘minor’ clause has something other than a verb as its head. e.g. How about a cup of tea? Oh for a horse! What a mess! Why not go home? To think you were there all the time!

modal auxiliary. E.g. will, might, must. Modal auxiliaries are important for communicating complex ideas because they express meanings such as possibility and obligation, but their distinctiveness lies in their grammar. They are auxiliary verbs which:

  • have no -s in the present tense even when their subject is third-person and singular: She can swim. (not: *She cans swim. Compare: She knows how to swim.)
  • are always either past or present tense, so they cannot occur in infinitival, imperative, subjunctive or participial clauses or in gerunds:
    It’s important to be able to swim. but not: *It’s important to can swim.
    Being able to swim is important. but not: *Canning swim is important.
  • select a following base form without to (exception: ought, which does license to)
    The main modal auxiliaries, by these criteria, are: will, would, can, could, may, might, shall, should, must and ought. More marginal members of the class are:
  • used (I usedn’t to swim or: I didn’t used to swim)
  • is/am/are/was/were licensing to (She was to regret it very soon)
  • dare (Dare you do it? or: Do you dare to do it?)
  • dared (Dared he do it?or: Did he dare to do it?)
  • need (Need you go so soon? or: Do you need to go so soon?)

modify, modifier, modification. If one constituent modifies another, the modifying constituent (the ‘modifier’) is subordinate to the modified one (the ‘head’), and is not ‘licensed by it. ‘Modifier’ is the name of a grammatical function, contrasting with ‘subject’ and ‘complement’, both of which are licensed. For example:

  • In sweet tea, sweet modifies tea so that the two words together mean ‘sweet tea’ instead of just ‘tea’, and sweet is a completely optional, unlicensed, addition.
  • In walks quickly, quickly modifies walks so that together they mean ‘walks quickly’ and not just ‘walks’, and once again, the modifier is unlicensed.
  • But in She ate it, neither the subject she nor the object it is a modifier, because she is licensed by the past-tense verb ate (which requires a subject) and it is licensed by the lexeme eat (which allows an object, though it does not require one).

The distinction between modifiers and other functions lies in their grammar, and not in their meaning. In terms of meaning, every subordinate constituent could be said to ‘modify’ the meaning of the head to change it towards the meaning of the whole phrase. So in She ran quickly, the subject she and the modifier quickly both change the basic meaning of ran to make it more precise.

A modifier positioned before the head is called a ‘premodifier’, and one positioned after the head is a ‘postmodifier’; for example, in big books about butterflies, the head is books, its premodifier is big and its postmodifier is about butterflies. A modifier of a verb is called an ‘adverbial‘.

mood. In some languages, finite verb inflections are classified as ‘indicative’, ‘imperative’ or ‘subjunctive’; this contrast is called ‘mood’. Indicative verbs are used in ordinary statements and questions in main clauses, with subjunctives used in special circumstance such as some subordinate clauses or in referring to hypothetical situations.

morpheme. A morpheme is the smallest unit of morphology, so it is the smallest part of a word that relates to the word’s grammar or meaning. A word may consist of
• one morpheme: house
• two morphemes: house+s, hous+ing, foot+ball
• three or more morphemes: house+keep+ing, un+happi+ness, foot+ball+er.
Morphemes may be classified as bases or affixes.

morphology. Morphology is the study of the internal make-up of words defined in terms of morphemes: one or more bases, which may be changed either by the addition of affixes, or in more complex ways such as vowel change, as in sing ~ sang. Morphology may be used in two ways:

  • inflectional morphology: to inflect a single lexeme house-s, walk-ed, foot ~ feet
  • lexical morphology: to relate distinct lexemes by:
    • derivation, linking single lexemes: teach-er, de-motiv-ate, rise ~ raise
    • compounding, linking one lexeme to two or more others: bookshop, blood-red

Derivatively, morphology may also refer to morphological structure, as when we say that irregular verbs have irregular morphology.

multi-clause sentence.See complex sentence.

negative, negation. See affirmative.

negative concord. In non-Standard English, a negative may be expressed more than once; for example:
We didn’t see nobody.
I do
n’t want to go nowhere.

I didn’t tell nobody nothing.

Nor never none / Shall mistress be of it, save I alone. (Shakespeare)
Traditionally, this construction is called ‘double negative’, but as the last two examples show, this name is misleading, and linguists prefer the term negative concord. Negative concord is not found in modern Standard English, though it was part of Shakespeare’s English. The equivalent Standard forms would use any- forms instead of the negatives, as in:
We didn’t see anybody.
I don’t want to go

I didn’t tell anybody anything.

Contrary to the claim often made in prescriptive grammar, negative concord is not illogical, because, as a matter of fact, an example such as We didn’t see nobody has exactly the same meaning as its Standard equivalent, We didn’t see anybody. The difference lies in the grammatical rules for signalling negation through the choice of pronouns and determiners: words like any in Standard grammar, and words like no in non-Standard.

Negative concord is obligatory in many standard languages such as Standard French (e.g. Marie ne dort pas, literally ‘Mary doesn’t sleep not’, translating Standard English Mary doesn’t sleep.), and was normal in earlier periods of English.

neologism. A neologism is a newly-created word, whether borrowed from another language, created out of ‘native’ material by derivation, or created from scratch (e.g. blurb)

neuter. See gender.

nominal, nominalize, nominalization. The term nominal relates to noun, so if a text uses a lot of nouns it may be described as ‘highly nominal’. Nominalization is a process which turns something other than a noun into a noun (or a clause into a noun phrase). For example, nominalization turns the verb arrive into the noun arrival, and red ino redness. Even more generally, it can be applied to any way of expressing a clausal meaning in a noun phrase, such as in gerunds (e.g. We had a holiday in Spain. compared with We’re looking forward to having a holiday in Spain.)

nominative. See case and pronoun

non-count noun. See count noun.

non-finite. See finite.

non-restrictive relative clause. See relative clause.

non-Standard English. See Standard English.

notation. Grammarians have developed a number of useful conventions for discussing grammar. They include:

  • * for ungrammatical sentences and ? where grammaticality is uncertain.
  • italics or underlining (or both) for words that are being discussed or quoted as examples, rather than used in the ordinary way.
  • small caps for lexemes.
  • Bracketed strings or trees for constituent structure.

noun. E.g. cat, person, arrival, purpose, Elizabeth, London, them, who. Nouns – the largest word class of all – are sometimes called ‘naming words’ because they name (more technically, refer to) people, places and things. This may be a useful way to remember what nouns are, but doesn’t always help to distinguish nouns from other word classes because these can also have similar meanings to nouns; for instance, places can be referred to by preposition phrases (behind the sofa, meaning ‘the place behind the sofa’) and actions by verbs (She arrived, stating her arrival). Moreover, the definition is circular because we can only recognise ‘things’ by the noun or noun phrase that refers to them; for instance, lightning must be a thing because we only have a noun for it, whereas thunder need not be a thing because we can use a verb (It thundered).

Nouns are subdivided into:

  • common nouns: dog, wine, time, day, teacher
  • proper nouns (‘names’ in the more conventional sense): Mary, London, Wednesday, Dad
  • pronouns: me, myself, who, whoever.
    The most reliable way to recognise nouns is by their grammatical behaviour. A noun can be used – either alone or as the head of a noun phrase with a specifier such as the – as the subject or object of a verb or of a preposition (e.g. The dog was hungry, Mice love cheese, She lives in Manchester.)

See also count noun, abstract noun, collective noun).

noun phrase. E.g. big books, the end of the road, the girl next door, someone else, those that I threw away, our Mary. A noun phrase is a phrase whose head is a noun, and which may also contain a determiner or genitive functioning as specifier (a book, my book, Mary’s book) as well as various kinds of modifier such as adjectives (a big book), nouns (a grammar book), preposition phrases (a book about grammar) or relative clauses (a book that I bought yesterday). The head noun may be missing through ellipsis (e.g. You could have the red tie, or would you prefer the blue?) One special kind of noun phrase is a free relative, and another is a gerund.

Noun phrases have a wide range of possible functions:

number. The inflectional contrast between singular and plural is called ‘number’. The term number can also be applied to words such as three, but to avoid confusion it is better to call these ‘numerals’.

numeral. Numerals (also called ‘numbers’) are words that denote a number, such as six. Basic numbers are called ‘cardinal numerals’, and are sometimes classified as determiners, in contrast with ‘ordinal numerals’ such as sixth, which are often classified as adjectives. However, both kinds of numeral can also be used on their own, like common nouns: I’ve had six. This is my sixth.

object. Object is the name of a grammatical function. A verb’s object, like its subject, can easily be identified by its grammar:

  • Class: it is normally a noun or a noun phrase:
    • Nobody understands Jane.
    • Nobody understands Jane’s behaviour.

(A restricted class of noun phrases such as last night can be used as adverbials as in I slept well last night; such examples are easy to distinguish from objects because they can be replaced by preposition phrases such as during the night.)

  • Word order: the object normally stands just after the verb (in contrast with the subject, which stands before the verb). So the normal order of elements in an English clause is: subject — verb — object (SVO).
  • Licensing: The object is a type of complement, so it is licensed by the verb; verbs that license objects are called ‘transitive‘. For example,
    • after devour, an object is required, so *I devoured. would be ungrammatical.
    • after eat, an object is possible but not obligatory, so we could say either We ate it. or: We ate.
    • after arrive, an object is not allowed, so *We arrived it would be ungrammatical.

Objects can be distinguished from subject complements by the following properties:

  • Most objects can become subject if one changes the voice from active to passive: Thomas prepared a lovely meal (object of active) – A lovely meal was prepared by Thomas (subject of passive). Subject complements, by contrast, can never become subject of a passive, as illustrated in the pair Their daughter became a teacher (subject complement of active) and *A teacher was become by their daughter (ungrammatical passive)
  • Subject complements can be adjectives, whereas objects cannot: He seemed sad (subject complement), but *He saw sad (object).
  • Semantically, subject complements describe the subject referent: in Jill became a teacher, ‘a teacher’ describes a property that Jill came to have; in contrast, objects normally have a different referent from the subject: in Jill met a teacher, the teacher is a separate person.

A verb‘s object often refers to the patient’, as in We ate the bread, where the bread is affected by the action of eating. But the object may have many other semantic roles:

  • We like bread. (The bread was not affected by the liking.)
  • We baked bread. (The bread was created by the baking.)

In all the examples given so far, the object is a ‘direct object’. In contrast, the pronoun them in We gave them a present is called an ‘indirect object, which typically occurs with a following direct object and identifies a person who receives the direct object’s referent or benefits from it. A noun or noun phrase functioning as indirect object can often be replaced by a preposition phrase headed by to or for:

  • We gave the children a present – We gave a present to the children.
  • We made the children a cake – We made a cake for the children.)

Grammarians often extend the term object from verbs to prepositions. In this usage, London is the object of from in the phrase from London. There are obvious similarities between the objects of verbs and of prepositions, including their position after the head, and the inflectional form of personal pronouns: accusative rather than nominative, giving with me rather than *with I (a similarity which is even more obvious in languages such as Latin and German where prepositions take the same range of cases as verbs).

object complement. Like a verb’s subject complement, its object complement adds information about the verb’s object. For instance, in It made Mary happy, the object complement happy describes the referent of the object, Mary. Like the subject complement, the object complement may also be a noun or noun phrase (e.g. They made her president)

objective case. See case.

ordinal numeral. See numeral.

orthography. (The study of) correct spelling.

parse, parsing. A traditional grammatical exercise for school-children was to ‘parse’ the words in a sentence by assigning each one to its word class or ‘part of speech’ (Latin: pars orationis), as well as giving a fixed range of information about its inflectional classification and about its function in the sentence. For example, in the previous sentence, the word words would be described as:

  • a common noun
  • the plural of word
  • the object of the verb parse.

part of speech. See word class.

particle. See phrasal verb.

participle, participial. Verbs in their ing-form and en-form (e.g. taking, taken) are traditionally called ‘participles’, though this term is sometimes reserved for particular uses of these forms, as explained below.

The contrast between ing-form and en-form is relevant to two important grammatical contrasts:

  • aspect, where the terms present (participle) and past (participle) are partly justified by the link between the perfect and events in the past:
    • ing-forms in progressives (She is taking some photos)
    • en-forms in perfects (She has taken some photos)
  • voice, where the term participle reflects important grammatical similarities between the clauses concerned, such as their ability to modify nouns or to function as adverbials.
    • ing-forms in actives
      • The girl taking the photos is Jane. (modifying girl)
      • Coming home from work I met a friend (adverbial)
    • en-forms in passives
      • The photos taken by Jane are these. (modifying photos)
      • Distracted by the buzzing fly, she started to think of that night’s party. (adverbial)

Ing-form participial clauses generally look just like gerunds, so some grammarians don’t distinguish them while others do.

passive. See voice.

past participle. See participle.

past tense. A past-tense verb (‘a verb in the past tense’, sometimes called ‘preterite’), like a present-tense verb, can be recognised by a combination of characteristics:

  • Morphology: a past inflectional form of the verb. In regular verbs it is formed by means of the suffix (e)d; the past tense of irregular verbs is formed in a variety of ways (e.g. sang, thought, put, did, was). In a few verbs such as put and hit, the past form is the same as the plain present one, so morphology alone isn’t always enough.
  • Syntax: occuring in finite clauses, which normally have a subject (e.g. She put it on, but not: It’s hard to put it on.
  • Meaning: locating the situation in past time: I went to bed. She put it on as she went out. She was naughty. She liked sausages.

Secondary uses include:

  • In backshift, the past tense follows that of the matrix clause, and the situation can be either present (I thought you still loved her) or future (You told me the exams started next week).
  • In remote conditional clauses it indicates improbability (I’d do it if you paid me) or counterfactuality (I’d do it if I had more time).
  • Other restricted sequences such as I wish I knew the answer. It’s time you were in bed.

Modal auxiliaries,can occur in the past tense without a past time meaning in main clauses:

  • with a remote conditional subordinate clause (If he did something like that again, I would/should/could/might report him),
  • just indicating politeness or tentativeness (Could you please open the window? She might be right after all).

patient. Patient is the name of a semantic role. In Mary kissed Bill, Mary (the person, not the word) is the agent and Bill is called the ‘patient’ (in the sense of ‘person affected’). The patient is often, but not necessarily, referred to by the syntactic object.

perfect. The perfect is formed by using the auxiliary verb have with an en-form verb e.g. has shown, had taken, have helped. have has almost the same range of inflectional forms as any other verb, so we can distinguish the present perfect (e.g. She has seen it. They have seen it.) from the past perfect (e.g. She had seen it ) and other uses such as Having seen the evidence, I’m convinced.

The present perfect has a different meaning from the other uses of the perfect. Like the simple past tense, the present perfect locates a situation in past time: compare She has gone to lunch and She went to lunch, which both locate the going to lunch in the past. This meaning can be described as locating the ‘event time’ before ‘now’, the present moment. The meanings, however, are different in two ways which justify the description ‘aspect’ rather than ‘tense’.

  • The present tense component of the present perfect links this construction to present time: it has in varying ways what is often called ‘current relevance’. For example, in She has lived in Brighton since 1995 we understand that she still continues to live there now; in She has gone to lunch we infer that she is still at lunch; in The Prime Minister has resigned (said on a news bulletin) the connection with the present is that this is ‘hot news’, something that has just happened, in contrast with The Prime Minister graduated from Oxford
  • The past time cannot normally be made specific by the use of a definite time adverbial such as yesterday; so although we can say Jill has visited us recently (‘at some recent time’), we cannot say *Jill has visited us two days ago.

In contrast, the past perfect introduces an additional point of time, the ‘reference time’, which is between the event time and the present; for example, in When I rang, he had already left, the reference time is the time of ringing, which is after the leaving and before now. Since the past perfect is the only way to distinguish the reference time from the event time, it makes no difference whether or not the event was relevant then or indefinite. Consequently, we can say I heard that the Prime Minister had graduated twenty years ago, with a definite time (twenty years ago) and without implying any continuing relevance at the reference time (the time of hearing). Much the same is true when the perfect is combined with other non-present forms of have, because once again the perfect is the only available way of expressing the meaning usually carried by the simple past. For example, They are believed to have left the country last week. means the same as: They left the country last week, it is believed. In all these constructions, in contrast with the present perfect, the perfect qualifies as a tense rather than an aspect.

Languages such as French and German often use a periphrastic construction similar in form to our perfect to translate the English simple past. For instance, I saw Mary yesterday would be J’ai vu Marie hier in French and Ich habe gestern Marie gesehen in German (in both cases literally ‘I have seen Mary yesterday’).

performative verb. See speech act.

periphrastic tense. See tense.

person. In grammar, a distinction is made between ‘first person’, ‘second person’ and ‘third person’ according to the person (or thing) referred to:

  • The first person is the speaker or writer, or a group containing the speaker.
  • The second person is the addressee or a group containing the addressee (but not the speaker)
  • everyone and everything else belongs to the ‘third’ person.

These persons are distinguished in the personal pronouns (I — you — she, etc.; myself — yourself – herself, etc.).
In some languages, a verb agrees with its subject in both person and number; this was true in earlier stages of English (Shakespeare’s I have — thou hast — he hath or he has) but agreement now survives in Standard English only in the present tense of most verbs (I/you/they walk — she walks), and in the past tense of the verb be (you/they were — I/he was).

personal pronoun. See pronoun.

phrasal verb. Phrasal verbs license a specific particle — a preposition without its complement, such as up, in, down.

  • He turned up late. (turned up = arrived)
  • We turned in for an early night. (turned in = went to bed)
  • She turned the offer down. (turned … down = rejected)

The examples show that the choice of particle can profoundly affect the verb ‘s meaning, so each of these examples of turn is in fact a distinct idiom, with its own distinct meaning. Such verb+particle combinations are often informal synonyms of more formal verbs; so give up = abandon, turn down = decline, fall out = disagree, and so on.

Particles are exceptions to the normal word-order rule that forbids other constituents to stand between a verb and its object, so She turned down the offer is possible as an alternative to She turned the offer down. This position between verb and object is available to particles even when the verb has its normal, non-idiomatic, meaning, as in She put down her book or She lifted down her case.

Phrasal verbs are similar to prepositional verbs, which also tend to be idiomatic; but their syntactic structures are quite different because the preposition of a prepositional verb is the head of an ordinary preposition phrase, so it cannot move after its object. Consequently, we find contrasts like the following:

  • phrasal verb: She looked up the word. Or: She looked the word up.
  • verb + preposition phrase: She looked up the street. But not: *She looked the street up.

The two kinds of verb are combined in phrasal-prepositional verbs such as put in put up with (where up is a particle and with is an ordinary preposition).

phrase. A phrase is a group of words containing one word – its head and all the constituents (words or phrases) that are subordinate to it. For example, in Small babies cry, the phrase small babies consists of the head, babies, and its modifier small. Similarly, the entire clause Small babies cry is a special kind of phrase (called a clause) consisting of cry and the smaller phrase subordinate to cry, its subject small babies. The way in which the words in a phrase are organised around the head word is called the phrase’s structure, and is described in terms of smaller constituents and their grammatical functions such as ‘head’ and ‘modifier’. (Exceptionally, the analysis of clause structure which recognises the auxiliaries and the lexical verb as a ‘verbal group’ allows this verbal group to be the clause’s head.)

Phrases allow us to express complex meanings which would not be expressible if we could only use single words; so by adding different modifiers to the word babies we can define different types of baby: sleeping babies, crying babies, hungry babies, babies with a lot of hair, babies who don’t sleep, and so on without limit. Each subordinate constituent contributes to this complex meaning in a different way, determined in complex ways by

  • its own meaning (e.g. small defines size whereas heavy defines weight)
  • its grammatical function (e.g. small babies is linked to different semantic roles when it has different functions, as in Small babies cry compared with I like small babies.)
  • the licensing properties of the head (e.g. small babies are the agent in Small babies drink, but not in: Small babies are lovable.)

Phrases can be classified according to the word class of their head word; so for example, small babies is a noun phrase because its head is a noun. Similarly, we have: • adjective phrases: She almost immediately ordered a very large ice-cream for every pupil.
adverb phrases: She almost immediately ordered a very large ice-cream for every pupil.
preposition phrases: She almost immediately ordered a very large ice-cream for every pupil.

determiner phrases: Almost every seat was taken.
The exception is the term verb phrase, which is best avoided because it is used in too many different, and conflicting, ways. The best name for a phrase headed by a verb is the well-established term clause.
This definition of phrases assumes that a phrase must always contain at least two words, so that in Babies cry, babies is not a phrase although small babies would be. You may come across a very different view of phrases which allows them to consist of nothing but their head word; in this view, babies in Babies cry would be both a word and a phrase — i.e. a noun and a noun phrase. This alternative view requires more complicated structures because of all the extra phrases that have to be recognised. On the other hand, it also allows simpler grammars because a single term such as noun phrase covers at least two possibilities, which are always available: a single unmodified noun, or a noun heading a many-word phrase. The definition given here is probably better for use in schools, where grammatical analysis is more important than grammar-writing.

In traditional grammar, the term phrase had a much narrower definition as either a (modern) preposition phrase or a non-finite clause.

pluperfect. In grammars of Latin, one of the inflected tenses of a verb was called the ‘pluperfect’, for example, amaveram meant ‘I had loved’. The term is now rarely used in English grammar, because had loved is normally called called the ‘past perfect’.

plural. A plural noun is an inflectional form which typically contrasts with a singular form; this contrast is called ‘number’. A plural noun typically refers to more than one example of whatever the noun denotes; for instance, whereas the singular dog refers to just one dog, its plural dogs refers to more than one dog.
In English, regular plurals contain the suffix -s or –es (e.g. cats, dogs, horses, wishes), but there are nouns with irregular morphology (e.g. mice, formulae). Some plural nouns have no corresponding singular:

  • oats is plural but has a meaning like that of a non-count singular such as wheat
  • scales is plural but has a meaning like that of the count singular balance.

polarity. Polarity is the contrast between affirmative and negative. See auxiliary verb, interrogative tag.

polysemy, polysemous. Most common words have a range of possible meanings rather than just one single meaning; for example, even an apparently straightforward noun such as book may refer to a physical object (I’ve lost that book) or to the abstract contents (I’ve finished writing that book). This phenomenon is called ‘polysemy’ and words like book are described as ‘polysemous’. (Contrast homonymy.)

positive. See affirmative and comparative.

possessive. See genitive.

postmodifier. See modify.

postposition. See preposition.

predicate. The predicate is that part of a clause which, according to one of the three main analyses of clause structure, is not the subject. So, in the sentence Clare visited her new school on Saturday, the subject is Clare and the predicate is visited her new school on Saturday.

predicative. See adjective.

prefix. See affix.

premodifier. See modify.

preposition. E.g. of, at, over, by, with. Like verbs, prepositions can license a following noun or noun phrase which is often called their object; but unlike verbs, they do not normally inflect
We got home at midnight.
Did you come here
by car?
Are you coming
with me?
They jumped
over a fence.
What’s the name
of this street?
I fell asleep
during the film.

With some prepositions, the object may be omitted:

She put her hat on her headShe put her hat on.

She went inside the house. — She went inside.

(Examples without an overt object would traditionally be classed as adverbs.) Prepositions that accompany phrasal verbs (e.g. give up) cannot have an object and are called ‘particles’. Some prepositions can also license objects other than nouns or noun phrases:

He came out from behind the curtain. (from + preposition phrase)

He came out of the room. (out + preposition phrase)

I regard that as outrageous. (as + adjective)

It went from bad to worse until recently. (from/to + adjective, until + adverb)

The traditional term preposition reflects the fact that these words are normally positioned before their object noun or noun phrase. Some languages regularly use postpositions, and a few of these are even found in English (e.g. a month ago, this letter notwithstanding).

Prepositions often indicate

  • time (at midnight/during the film/on Friday),
  • position (at the station/in a field)
  • direction (to the station/over a fence).
    But there are many other meanings, including
  • source (from the cow)
  • means (by car)
  • accompaniment (with me).
  • apposition (City of London)

In interrogatives and a few other structures such as relative clauses, prepositions are often ‘stranded’ (i.e., left without a following object after the verb) because their object has been removed or fronted:
Who did you go out with?
I found the book
t I was looking for.
In more formal style we avoid stranding by putting the preposition before whom or which (with whom, about which etc):
With whom do you wish to speak?

His father, for whom he has enormous respect, has just turned 90.

prepositional passive. See voice.

prepositional verb. A prepositional verb licenses a particular preposition when it has a particular idiomatic meaning; for example, when look means ‘care for’, it licenses after (as in She looked after him). Prepositional verbs can look like phrasal verbs, which can also have idiomatic meanings, but they license quite different syntactic structures

preposition phrase. E.g. in my house, to the seaside, of Mary. A preposition phrase (often called a ‘prepositional phrase’) is a phrase whose head is a preposition. The most important functions of preposition phrases are:

Within a preposition phrase, the preposition is usually followed by its object (though this may often be omitted) and may be preceded by a premodifier:

  • preposition + object: The same thing had happened [before their wedding].
  • premodifier + preposition + object: The same thing had happened [many years before their wedding]
  • preposition: The same thing had happened [before].
  • premodifier + preposition: The same thing had happened [many years before].

prescriptive grammar. See grammar.

present participle. See participle.

present tense. A present-tense verb, like a past-tense verb, can be recognised from a combination of properties:

  • morphology: either a plain present or s-form inflectional form, as required by the subject (e.g. They walkbut: She walks); in regular verbs, the plain present form is the same as the base form, but in the verb be the two are distinct (are versus be).
  • syntax: occurring in a finite clause, normally with a subject (e.g. They walk to work but not: It’s nice to walk to work.)
  • meaning: denoting a situation located in present time, including the deictic ‘now’. How the time of the situation relates to ‘now’ varies according to the type of speech act and the type of situation:
    • In a running commentary, the time of each event reported is (roughly) the same as the current ‘now’ (e.g. Smith passes the ball to Brown, and Brown kicks it into touch)
    • In a performative, the event reported is the utterance itself, so the two times are the same (e.g. I promise to come home early)
    • If the situation is a state, its time includes ‘now’ (e.g. My parents live in Nottingham. Jill is extremely bright.). This possibility includes the present be when used in the present progressive, in which the time of the situation includes ‘now’ (He’s mowing the lawn; I’m living in my sister’s place just at the moment)
    • If it is a habitual or characteristic recurrence of an event, it is the habit or recurrence that includes ‘now’ (I generally watch TV in the evenings, The Committee meets once a fortnight).

The present tense also has secondary uses:

  • for an event which is expected to happen in the future (I leave tomorrow. The train arrives in three hours).
  • in a subordinate clause whose matrix clause denotes an event in the future (Send me a card when you arrive! We may not be able to feed everyone who comes.)

preterite. See ‘past tense’.

progressive. The progressive aspect (also called ‘continuous’) is normally formed by combining a present participle with be, as in is working, were trying. The primary use of the progressive is to present the event or situation as being in progress at the time in question:

His health was deteriorating rapidly when I saw him.

They are still working even though they retired years ago.

She must be walking very slowly.

In a secondary use the progressive presents a future event as already planned:

I’m seeing my doctor tomorrow morning The progressive can also be combined with the perfect (e.g. He has been reading).

pronoun. E.g. me, him, he, his, himself, who, someone. A pronoun is a special kind of noun. for use in communicative situations where it is preferable to a common or proper noun or noun phrase. Traditionally, ‘pronoun’ and ‘noun’ were distinct word classes, and pronouns were described as being used ‘instead of a noun’ (hence the name pro-noun), but this glossary treats pronouns as a sub-class of noun, alongside common nouns and proper nouns. The main reason for this is that pronouns occur in the same range of functions as common and proper nouns and noun phrases: as subject, object of a verb, object of a preposition, and so on. However, the pronouns also differ from common and proper nouns in not normally taking specifiers (so we can’t say *the you or *a who).

The possibility of using a pronoun instead of a common or proper noun is important in a number of situations:

  • where a common noun or proper noun is dispreferred because the referent is the speaker or the addressee (e.g. I love you.) These cases normally require a definite personal pronoun used deictically.

where a common noun would be repetitive or pointless because the referent has already been identified (e.g. The girl said she was tired — not: The girl said the girl was tired.) These cases require a definite pronoun used anaphorically.

  • where the pronoun is needed to indicate some particular grammatical construction (e.g. the girl who came in …. or: Who came in?) These cases require a grammatically specialised pronoun such as a relative pronoun or interrogative pronoun.
  • where the pronoun allows a convenient combination of generality and openness about its referent (e.g. I’m looking for someone to do this). These cases require an indefinite compound pronoun.

Pronouns fall into the following classes and sub-classes:

  • definite
    • personal (‘to do with grammatical person’ — not ‘for people’): seven lexemes with various inflectional forms: I, you, he, she, it, we, they. We also include one (as in One does one’s best.) because it has a reflexive form (though it is less obviously definite). Some of the following differences are sometimes explained as distinct cases.
      • ‘accusative’: for general use except as listed below: me, you, him, her, it, us, them; one
      • ‘nominative’: for use as subject in a finite clause: I, you, he, she, it, we, they; one (some speakers also use these forms coordinations such as between you and I, as well as in a few other constructions such as It was I.)
      • genitive‘, for use
        • as a specifier: my, your, his, her, its, our, their; one’s
        • otherwise: mine, yours, his, hers, its, ours, theirs
      • ‘reflexive’: myself, yourself, himself, herself, itself, ourselves, yourselves, themselves; oneself
    • reciprocal: each other, one another
    • demonstrative: this/these, that/those
    • relative: who/whom, what, which, whose, when, where; and on some accounts, that; also whoever, whatever, etc.. These words are sometimes called wh-pronouns.
  • indefinite
    • quantifiers some, any, none, each, all, both, either, neither
    • compound (someone, no-one, anyone, everyone; somebody, etc.; something, etc.; somewhere)
    • interrogative (who, what, which, whose; also: whoever, whatever, etc., and possibly when and where ) These words are also sometimes called WH-pronouns.
    • dummy there (as in There was a fly in my soup, wasn’t there.)

In many cases, this classification of pronouns matches the classification of determiners; indeed, almost every determiner is matched by a pronoun.

proper noun. A proper noun is a noun which is used as a name, such as Mary, Britain, Wednesday. It refers to an individual person or thing so it is a count noun and normally singular, but unlike singular count common nouns, it either needs no specifier at all (e.g. Mary, Britain) or always combines with the (e.g. The Sudan, The Thames). Some nouns can be either common or proper; for example, mother is a common noun in Her mother is tall. but a proper noun in I’m going to tell Mother. Proper nouns are spelt with a capital letter, but the capital letters also apply to modifiers and determiners (as in New York, The Andes), where the whole noun phrase functions as a name Proper and common nouns are sometimes hard to distinguish when used with the; e.g. the queen, the pope. Moreover, proper nouns are often converted into common nouns, as in Let’s listen to some Beethoven.

punctuation. Punctuation includes any conventional features of written presentation other than spelling, so it includes:

  • the standard punctuation marks (. , ; : ? ! — — ( ) “ ” ‘’ )
  • word-spaces and paragraph breaks
  • capital letters and other special fonts such as bold, underlined and italic
  • apostrophes
  • bullet points.

Punctuation has a number of different roles which sometimes conflict with each other:

Used as an indicator of speech acts, punctuation applies to main clauses rather than to whole sentences, so when a sentence contains more than one main clause it may require different punctuation for each one; but this necessarily conflicts with the idea that ‘sentence punctuation’ marks the boundaries of a syntactic sentence:

  • Did you see that back flip? And she’s only eight years old!
  • Can that really be true? But I agree that it’s possible.
  • What a rotten idea! Or am I missing something?

Used as a guide to information, punctuation is a resource that experienced writers can exploit for deliberate effects, in ways that go against the purely syntactic structure, either by dividing syntactic sentences into smaller chunks (e.g. We are prepared to strike. Now. For our rights. Because we have been exploited for long enough. And because no other course of action is open to us.) or by grouping syntactic sentences into larger units. This happens particularly often with semi-colons (as here); dashes can be used in the same way, especially in less formal writing – they are common in personal letters and emails.

However, one punctuation pattern is generally considered wrong: the ‘comma splice’.

quantifier. Quantifiers include indefinite determiners or pronouns such as some, any, none. which indicate some proportion of a whole class; e.g. some children refers to a subset of children smaller than the whole but larger than a single child. The term is also used more broadly to include other kinds of word with similar meanings, such as many and all.

question. See speech act.

question word. See interrogative word.

refer, referent. A constituent is said to ‘refer to’ the people, things, events, places and so on that it picks out (whether actual or imaginary, such as Father Christmas). The entity referred to by a word is called its ‘referent’. For instance, in the sentence The birds woke me this morning,

  • me refers to the person speaking or writing.
  • this morning refers to some time in the morning of the day when the speaking or writing occurred.
  • the birds refers to the birds in question.

Unlike the constituent’s denotation and connotation, its referent is not part of its semantics

reduced relative clause. See relative clause and participle.

reflexive pronoun. See pronoun.

register. The language that we use varies with the situation in which we use it: formal or informal, technical or lay, written or spoken, and so on. This kind of variation is described in terms of different ‘registers’ such as ‘legal English’ or ‘the English of casual conversation’ (in contrast with variation between groups of users, which are called ‘dialects’).

regular. When words are described as ‘regular’ or ‘irregular’, this normally refers specifically to their inflectional morphology. For example, English regular verbs form their past tense by adding –ed, as in walk – walked, but irregular verbs use different changes such as take – took, buy – bought or sing – sang. Similarly, mouse is described as an ‘irregular noun’ because of its plural mice. This rather specialised use of the terms regular and irregular should not obscure the fact that virtually any rule may have exceptions (which are therefore ‘irregular’), including the rules of syntax; for example, although typical degree modifiers such as sufficiently stand before the word they modify (as in sufficiently big), the word enough, exceptionally, stands after it (as in big enough).

relative clause. A relative clause is a subordinate clause that typically modifies a noun. For instance, in a cake which he made yesterday the underlined relative clause adds further information about the cake: that he made it yesterday. The relative pronoun which has an anaphoric link to cake, its antecedent, while at the same time showing that the relative clause is grammatically subordinate to this antecedent.

Although such examples are typical, there is a wide range of different ways of both using and forming relative clauses:

  • The clause may be ‘restrictive’ or ‘non-restrictive’ (sometimes called ‘defining’ and ‘non-defining’), according to whether it helps to restrict the noun’s meaning by making it more precise, or adds further information about the referent of the noun. The non-restrictive meaning is generally signalled in writing by a comma.
    • The student who was taking five subjects was finding life very difficult. (Which student? The one taking ….)
    • Sophy, who was taking five subjects, was finding life very difficult (And incidentally, they were taking …)
  • The antecedent of a non-restrictive clause may be a noun phrase, a preposition phrase or a clause. These options are only available for non-restrictive relative clauses:
    • All the other students, who had worked more slowly, had to stay behind to finish off. (who = all the other students)
    • We met behind the gym, which was a funny place to meet. (which = ‘behind the gym’)
    • We met behind the gym, which made me suspicious. (which = ‘we met behind the gym’)
  • The anaphoric element may be a relative pronoun (which, who, whose, where, when, why) or it can be implicit, with or without that to mark the start of the relative clause. (The word class of that is a matter of dispute.) The latter option is generally only available for restrictive relative clauses, but it is a common form in informal speech.
    • We ate a cake I made yesterday.
    • We ate a cake that I made yesterday.
  • In a ‘reduced relative clause’ the relative clause is not finite. It may be either participial or infinitival (with to, possibly accompanied by for introducing the subject):
    • We need a receptacle consisting of two parts.
    • We need a receptacle disguised as a bowl of fruit.
    • We need a receptacle in which to store it.
    • We need a receptacle for users to store it in.

In an example like I paid for what I bought. the phrase what I bought is often called a ‘free relative clause’.

relative pronoun. See pronoun, determiner, relative clause.

reported speech. See direct speech.

restrictive relative clause. See relative clause.

role. See semantic role.

root. A word’s etymological origin is often called its ‘root’; for example, the Greek word σχολή (skhole) meaning ‘leisure’ is the root of our word school.

root word. See base.

semantic, semantics. When we give the dictionary meaning of dog, we are defining its semantics — the meaning that it carries in the language system. Similarly for any other constituent: its semantics is the meaning that it carries with it (notably its denotation and connotation), in contrast with the meaning that is provided from the context, including its referent (e.g. the particular dog referred to on a particular occasion).

semantic role. A clause describes some kind of situation — an event or a state — in which various people or things are involved, each with a different ‘role’. The roles can be distinguished in terms of very general labels such as ‘agent’, ‘patient’ and ‘instrument’, which in turn can be related to syntactic functions such as ‘subject’ and ‘object’. The relations between semantic roles and syntactic functions are complex and vary from verb to verb; for example, the subject of a verb such as break can identify the agent (She broke it with the hammer), the patient (It broke) or the instrument (The hammer broke it)..

sentence. The sentence is the largest unit of grammar, i.e. a string of words held together by syntactic relations of either subordination or coordination which is not part of an even longer such string. Syntactic relations are essential to the definition of a sentence, in contrast with a text which may consist of more than one sentence and be united by mere coherence rather than by syntax

Sentences are important in writing because they are marked by punctuation. In straightforward cases, boundaries between sentences (as defined above) are marked by ‘sentence punctuation’ — a capital letter at the beginning, and a full stop, a question mark or an exclamation mark at the end. A particularly common misuse of punctuation is the comma splice. For the classification of sentences as ‘single-clause’ or ‘multi-clause’, rather than as ‘simple’, ‘compound’ or ‘complex’, see complex sentence.

sentence type. See clause type.

short passive. See voice.

simple sentence. See complex sentence.

single-clause sentence. See complex sentence.

singular. See number.

situation. A verb is often said to denote a ‘situation’, to avoid more specific terms such as action, event or state. This terminology allows us to talk about the meanings of verbs as different as jump, fall, die, think, know, exist and happen.

specifier. Within a noun phrase such as the book or the teacher’s book, the head is the common noun book, but there is no widely accepted name for the function of the first part (the determiner the or the genitive the teacher’s). However, the term ‘specifier’ is sometimes used in this sense, and is used throughout this glossary. Specifiers can be recognised by two criteria:

  • A singular count common noun such as hammer or number (but not a non-count noun such as coffee or beauty, or a proper noun such as Mary) normally requires a specifier, so we can say with the hammer or with Mary’s hammer but not: *with hammer.
  • Only one specifier is possible per noun phrase, so (adopting a narrow definition of ‘determiner’ that excludes words such as all and two) we cannot combine two determiners (e.g. *the this book) or a determiner and a genitive (e.g. *a my friend).

speech act. A main clause is produced with some kind of communicative intention such as stating, questioning, requesting, inviting, ordering, begging and so on. These intentions distinguish categories called ‘speech acts’ (a term which applies equally to written or spoken texts, in spite of the word speech); so in using a main clause we are performing some kind of speech act.

The speech acts for which a clause may be used are tied to its clause type, but not in a simple way because a single clause type may be used for different speech acts and conversely a single speech act may be carried out using a variety of clause types:

  • Declarative main clauses are characteristically used to make statements: It is raining
  • Interrogative main clauses are characteristically used to ask questions: Are you ready?
  • Imperative main clauses are characteristically used to issue directives: Sit down! (‘Directive’ is a technical term covering acts of ordering, requesting, advising, permitting someone to do something, and so on.)
  • Exclamative main clauses are characteristically used to make exclamatory statements (a kind of exclamation): What a mess we’re in!

But other pairings are possible, for example:

  • The interrogative Could you pass me the salt? would normally by understood as a request (“Please pass me the salt”) rather than a question about your ability to pass me the salt.
  • The imperative Sleep well would normally be used to express a hope.
  • The declarative I promise to be home by six would normally be used to make a promise rather than to make a statement. promise is a ‘performative’ verb (like congratulate, apologise and beseech) which is used to perform the action it describes; so the above example is not simply a description of a promise, but is itself that promise.
  • Declarative clauses like He’s ready can be used to ask a question (“Is he ready?”) if spoken with rising intonation or punctuated with a question mark, and have complex meanings when combined with interrogative tags (He’s ready, isn’t he?).

As in so many areas of grammar, the grammatical classes, here clause types, are different from those of meaning, here speech acts.

speech marks. See direct speech.

split infinitive. See infinitive.

Standard English. Standard English is the variety of English used in public communication, particularly in writing. It is the form taught in schools and used in public situations. Whereas non-Standard English varies considerably from region to region within each English-speaking country, Standard English does not.
Standard English and can be spoken with any accent, regional or not.; so it is identified by its vocabulary and grammar; and even in grammar there are relatively few differences between Standard English and non-Standard English. For example:
We were given those bottles. (Standard English)
We was given them bottles. (non-Standard English)
These examples show Standard and non-Standard agreement. A different non-Standard grammatical feature is negative concord. Contrary to the claims of prescriptive grammar, non-standard varieties are not grammatically inferior to standard varieties; they simply conform to different rules.
Of course, Standard British English is not the only standard variety of English; other English-speaking countries, such as the United States, Jamaica and Australia, have slightly different standard forms. Moreover, because Standard English is the native variety of some speakers, it includes a wide range of registers, from highly formal to casually informal, and including speech as well as writing.

statement. See speech act.

stem. See base .

stranded preposition. See preposition.

structure. When constituents such as words and phrases combine grammatically they form a structure which can be defined in terms of the constituents, their functions, their classes and their order; for example, in the noun phrase the very big book we can identify:

subject. The subject is one of the four main functions in clause structure, distinct from the verbal head, complements (including objects) and adverbials. It is usually a noun or noun phrase (Kim has arrived; The boss wants to see you), though subordinate clauses are not uncommon (Inviting her husband was a mistake).

The following grammatical tests almost always point to the same result.

  • Class: The subject is a noun, noun phrase or content clause (e.g. Jane/Jane’s behaviour/That Jane was late upset us).
  • Word order: The subject is normally the last such constituent before the verb (in contrast with the object, which is the first one after the verb); e.g. John kicked it. (S V O) (Exception: examples like Here comes Mary or In the corner of the field stands an old oak tree, where the subject follows the verb.)
  • Word order: It is the position of the subject, before the verb or after the auxiliary verb, that typically distinguishes main-clause declaratives from interrogatives (e.g. declarative John is happy. and interrogative Is John happy?) (Exception: when the subject is itself an interrogative word or phrase, as in What happened?)
  • Optionality and finiteness: An English finite clause generally needs an overt subject, unlike languages like Spanish where the subject may be omitted (e.g. vengo or io vengo, ‘I come’); the only exception in English is for imperatives, where the subject you can be omitted, and often is. In non-finite clauses, the subject is generally implicit.
  • Agreement: In the present tense, present-tense verbs generally agree with the subject (she walks — they walk) but they never agree with the object; and in the verb be, the past tense agrees as well (she was — they were). For example, I must be the subject in Who am I?, whereas who is the subject in Who is in charge? (Exception: modal auxiliaries show no agreement with the subject.)
  • Pronoun form: A few pronouns have a special ‘nominative’ form for use as the subject: I saw her but: She saw me.

In contrast with these grammatical criteria for recognising subjects, meaning is a very poor guide because the subject’s referent may have a variety of roles in the meaning:

  • agent: Jane opened the door.
  • instrument: The key opened the door.
  • patient: The door opened. or: The door was opened.

subject-auxiliary inversion. In Have you finished?, the subject you and the auxiliary verb have are ‘inverted’, i.e. their normal order has been reversed. This construction is found in:

  • most main-clause interrogatives (Have you finished? What have you finished? — but not What happened?)
  • some conditional clauses (Had I known you were coming I’d have stayed at home)
  • after a fronted so or a negative fronted constituent (… and so have you. Not only have you made a mistake, but … Nowhere else would you experience such freedom.)

subject complement. In Mary seems a nice person, the phrase a nice person is a description of Mary, the referent of the subject, whereas in Mary met a nice person, this phrase is the object, and refers to a separate person from Mary. In the first example, a nice person is called a ‘subject complement’ because it is a complement and describes the referent of the subject. (Contrast object complements such as angry in She made him angry, which describe the object’s referent.)

Just like an object, a subject complement may be a noun or noun phrase, as above, but it may also be:

  • an adjective (Mary seems nice).
  • a preposition phrase (Mary is in a good mood).

Consequently, a good test to distinguish subject complements from objects is the possibility of replacement by an adjective; for instance, a nice person can be replaced by nice in Mary seems …, but not in Mary met …..

Verbs that allow a subject complement are sometimes called ‘linking verbs’.

subjunctive. In some languages, verbs have a special set of inflectional forms, varying in person, number and tense, called the ‘subjunctive mood’ which is typically used (especially in some subordinate clauses) to show that the clause expresses something other than a statement of what is known or certain, for which the indicative is used. For example, French subjunctive verbs are used after certain subordinators and in certain kinds of subordinate clauses, to emphasise the uncertainty already expressed in the main clause (e.g. after ‘I’m not sure that …’) or to present a situation as desirable rather than factual; and German subjunctives are used to report what others have said.
Although earlier stages of English had a full subjunctive in this sense, the subjunctive only survives in Modern English in fixed phrases such as So be it, Be that as it may or God save the Queen, and two very restricted constructions found in some formal varieties:

  • the base-form subjunctive has the verb’s base form even when an -s-form or should would otherwise be expected:
    The school requires that every pupil give (instead of: gives or: should give) a presentation.
    The proposal is that the Headmistress be (instead of: is or: should be) the chair.
  • the were-subjunctive which is sometimes used in conditional clauses instead of was:
    If I were you, …

subordinate, subordinator, subordination. Most constituents in a sentence are linked in the unequal relation of subordination (also known as ‘dependency’, so the subordinate constituent is a ‘dependent’), rather than the equality of coordination. For example,

  • a modifier is subordinate to the word it modifies: small boy (small is subordinate to boy)
  • a verb’s subject and complements are all subordinate to the verb: She made him happy. (she, him, happy are all subordinate to made, the head of the clause)

a preposition’s object is subordinate to the preposition, and the resulting preposition phrase is in turn subordinate to some other constituent: She lives in London. (London is subordinate to in, and in London is subordinate to lives.)

In each case, the subordinate constituent combines closely with the other word to make a phrase, in which the other word is the head:

  • The small boy made her happy.(The small boy is a phrase in which the head is boy, modified by small.)
  • The small boy made her happy. (The small boy made her happy is a phrase, and more precisely, a clause, whose head is made.)
  • The subordinate constituent generally combines with the head word’s meaning to define a more precise meaning: small boy is more precise than boy (because a small boy is a particular kind of boy), and he made her happy is more precise than just made.

The subordination may be signalled in various ways, including by means of:

  • a subordinator (e.g. if, that): I’ll help you if you want
  • a relative or interrogative pronoun (e.g. who): People who live in glass houses shouldn’t throw stones
  • a preposition: A picture of Mary stood on his desk.
  • the use of a non-finite (e.g. walking): I saw him walking home.

subordinate clause. A subordinate clause is subordinate to some constituent, with various possible grammatical functions including the following:

  • modifying a noun: The man who came to supper knew my father.
  • as an adverbial: She fell down when she went out.
  • as a verb’s subject: Eating toffee apples is fun.
  • as a verb’s complement: He said that he was ready.

A subordinate clause is part of a larger clause, called its ‘matrix clause’, so in each of the above examples the underlined clause is part of the clause that constitutes the entire sentence.

Subordinate clauses with different functions often have different structural markers as well, usually located at or near the start of the clause::


  • Some subordinate clauses have no marking. (I know you’re angry. The book I bought is on the table.)
  • Clauses that are quoted as direct speech are not subordinate clauses, but independent texts (She shouted, ‘Go away! Can’t you see I’m sleeping?)

Subordinate clauses may be subclassified into the following general types:

  • content clauses.
    • I see that it’s raining.
    • Whether it will ever stop is unclear.
  • comparative clauses
    • It cost more than I expected.
    • I want as much as I deserve.
  • adverbial clauses
    • I’ll come when I’m ready.
    • Although it was rather late, we went for a walk.
  • relative clauses
    • The woman who lives next door gave me some flowers.
    • Athens, which is the capital of Greece, is a fascinating place.

suffix. See affix.

superlative. See grade.

SVO (subject — verb — object). See clause-structure.

synonym, synonymous. Two words are synonyms if their meanings are the same or very similar; e.g. try – attempt; close — shut.

syntax, syntactic. Syntax is the part of grammar which is concerned with structure of sentences, i.e. with how words are organised in a sentence. It includes word order, phrases, subordination and agreement. (Contrast morphology.)

tag question. See interrogative tag.

tense. Tense is the name for the contrast between inflectional forms of a verb which indicate distinct times, such as are and were or walks and walked.– Each of these forms is called ‘a tense’. However, the term is used in two different ways, according to whether the contrasts are limited to a single verb (‘simple tense’) or whether combinations of verbs are also recognised as tenses (‘compound tense‘ or ‘periphrastic tense’).

1. A simple tense contrast is signalled by inflection and is primarily used to indicate differences of time. English verbs have two inflected tenses:
present, e.g. is, waits, run
past e.g. was, waited, ran
but unlike languages such as French, English has no simple future tense (although it does of course allow us to talk about the future).

The present-past contrast is primarily used to locate the time of a situation in present time (a period which includes ‘now’, the moment of utterance) or past time (time preceding ‘now’). Tense is thus a deictic category, with the interpretation dependent on the time of speaking or writing . (However, the tenses also have other non-deictic uses as explained in the entries for present and past.)

2. English also has periphrastic word-combinations that are sometimes called ‘tenses’:
perfect, e.g. has waited, had waited
progressive, e.g. is waiting, was waiting
These periphrastic tenses combine with the simple tenses and with each other to define eight ‘tense-aspect’ combinations:

  • present simple: waits
  • present perfect: has waited
  • present progressive: is waiting
  • present perfect progressive: has been waiting
  • past simple: waited
  • past perfect: had waited
  • past progressive: was waiting
  • past perfect progressive: had been waiting

Grammarians agree in recognising the perfect and progressive, though they disagree about whether to call them ‘tenses’ or ‘aspects’; this question arises in particular for the perfect. They also disagree, as explained under ‘clause structure’, about how a sequence like has been waiting fits into the overall structure of the clause.

Much more controversially, however, some grammarians also recognise a periphrastic future tense signalled by the modal auxiliary will (as in will wait, will have been waiting). Here the disagreement goes beyond mere terminology, as explained in the entry on ‘future tense’.

tensed. See finite.

text.  See discourse.

token. See word.

transitive verb. Verbs are traditionally classified as ‘transitive’ if they license a direct object and as ‘intransitive’ if they don’t; so destroy would be transitive and arrive intransitive. However, most verbs in English can be used either with or without a direct object:

  • roll: They rolled the stone back — The stone rolled back.
  • wash: I washed my shirt — My shirt washes well.
  • eat: I ate a banana — I ate noisily.

Moreover, direct objects are only one of the many kinds of complement that a verb may license, so the transitive/intransitive contrast has limited value on its own as a way of describing licensing possibilities.

The term transitive derives etymologically from the ‘transition’ of activity from the subject to the object, as in John kicked the ball where the kicking is taken to pass from John to the ball. However, in practice we use the term transitive for verbs that take a direct object even where such a transition doesn’t exist: John saw the ball; John imagined the ball

type. See word.

understood subject. A subject is described as ‘understood’ if it is clear from the context, but not actually expressed as a separate word. For example, in the imperative Come in! the missing subject is ‘you’, and in I want __ to go home, the understood subject of to go home (represented by the underscore) is the same as the explicit subject of want, i.e. ‘I’. In the simplest cases, the understood subject in a subordinate clause is the same as that of the main clause. However, the antecedent of the understood subject need not be the main clause’s subject; it may be:

  • the main clause’s object (I persuaded her __ to accept the job)
  • part of an adverbial (I sent it to Mary __ to read on the journey)
  • not part of the main clause at all (__ Inviting her husband was a mistake.)

This flexibility in the choice of antecedent creates some uncertainty, as in I met her coming home from work, where the subject of coming could be understood as either ‘I’ or ‘her’. Prescriptive grammar tries to reduce this uncertainty by allowing only subjects as antecedents. Other possibilities are called ‘dangling participles’, but ordinary speech provides many exceptions such as __ Coming home from work my hat blew off. The same issues arise with infinitival clauses such as These books are__ to read on the journey.

ungrammatical. See grammatical.

variety. No language is homogeneous, and variation is inherent in language. One way to discuss this variation is in terms of large-scale ‘varieties’ which combine a set of distinctive linguistic features with various kinds of external features:

  • place differences distinguish geographical dialects
  • social-group differences distinguish social dialects (or ‘sociolects’)
  • time differences distinguish different historical periods of the language.
  • situational differences distinguish different registers (e.g. formality differences).

verb. E.g. take, arrive, imagine, rain, be are all classified as verbs The best way to recognise a verb is by its ability to have a tense and a subject. For instance, arrive must be a verb because we can contrast present-tense arrives or arrive with past-tense arrived, and we can use both of these with a subject such as she, as in She arrived. In contrast, arrival cannot be a verb because it doesn’t have a tense and can’t be used with a subject, so we can’t say *She arrivals/arrivalled.

Verbs are sometimes called ‘doing words’ because they often name an action that someone does. This may give some initial idea of the kind of words included in the verb class, but, like other meaning-based definitions it does not provide a criterion for identifying verbs in a text. This is because a great many verbs do not describe actions: in particular many describe states, as in She believes in God, He has red hair, The soup is cold. This is why we need a more general term for a verb’s meaning, such as situation. Moreover, we can also use nouns instead of verbs for actions and states, as in the assassination of the president, her belief in God, etc.; indeed, even the most general terms for a verb’s meaning, including situation, are nouns.

A verb lexeme such as take has six inflectional forms, more than any other English word class, and verbs are the head of the most important syntactic unit, the clause. with a particularly important role of licensing complements. Moreover, they are deeply involved in some of the most important grammatical contrasts such as tense, aspect, voice, polarity and clause type, so they are arguably the most important word class of all. This primacy is reflected in the term verb, whose Latin source means simply ‘word’. A particularly important kind of verb is the auxiliary verb.

verbal group. In some analyses of clause structure, a lexical verb and its auxiliary verbs constitute a ‘verbal group’ or ‘verb phrase’; for example, has been writing would be a verbal group in He has been writing a book..

verb phrase. See clause structure, verbal group. In some analyses, the verb phrase is a very important syntactic unit. In this glossary we avoid using this term, because it is used in different ways by various grammarians, and could cause confusion

vocabulary. See lexicon.

voice. This is the traditional name for the contrast between ‘active’ and ‘passive’. ‘Active’ is the default so a clause is active if it does not have the distinctive passive properties. The following is a typical active-passive pair:

  • Fido bit Ben (active)
  • Ben was bitten by Fido (passive)

These have the same meaning, but present the information differently, according to whether they say something primarily about Fido or about Ben.

There are two main grammatical differences:

  • The nouns Fido and Ben are aligned differently with the grammatical functions in the clause: Ben is object of the active but subject of the passive; and Fido is subject of the active but object of the preposition by in the passive (called ‘the by phrase’).
  • Passives also have different verbs: a form of be (here was) followed by the form bitten, the en-form of bite, instead of bit, its past-tense form.

The terms ‘active’ and ‘passive’ reflect the fact that, in clauses expressing actions, the subject refers in the active to the agent , which ‘actively’ performs the action, but in the passive to the patient, which undergoes the action ‘passively’. But the voice contrast applies even to situations where there is no activity or passivity in the meaning; for instance, knowing someone is not an action, but the active/passive labels are applied to examples like Everyone knows Ben (active) and Ben is known by everyone (passive). Conversely, a ‘passive’ meaning as in My grandmother underwent surgery can be expressed in an active clause, in contrast with its passive Surgery was undergone by my grandmother.

Other kinds of passive clause differ from these typical examples in one or more of the following ways:

  • The by-phrase corresponding to the subject of the active is often omitted: Ben was bitten. In this case the clause is called a ‘short passive’.
  • Under certain conditions the verb get may be used instead of be: Ben got bitten (by the dog).
  • The passive clause may be participial, without a preceding be or get:
    • The boy bitten by the dog was rushed to hospital.
    • I saw it taken away or She had her son examined by a specialist.
  • Under certain conditions the subject of the passive clause may correspond to the object of a preposition in the active counterpart, rather than to the object of the verb. Such ‘prepositional passives’ are common in English, but are not possible in languages such as French and German.
  • Ben was barked at by Fido. (passive)
  • Fido barked at Ben. (active)

wh-cleft. See cleft sentence.

wh-interrogative. See clause type.

wh-word. See pronoun.

word. A word is a unit of grammar that can be selected and moved around relatively independently of other such units. In punctuation, words are normally separated by word spaces, but this is omitted in abbreviations (e.g. contractions such as I’m).
Words are the basic building blocks of grammar, so it is fortunate that they are generally easy to identify. Every time we write a word space we are recognising the boundary between two words, and our spelling identifies words by more abstract features than their pronunciation; for instance, if we write fair rather than fare, we are recognising two different but homophonous words.

But there are complications. For example, there is an uncertain boundary between compound words and two-word combinations of a modifier plus head:, is brick-red one word or two? And should we write land owner or land-owner (or even landowner)? Dictionaries give guidance, though not always the same guidance.

Another issue is that what we call ‘words’ in ordinary usage may be either ‘word forms’ or ‘dictionary words’ (technically calledlexemes’); for example, book and books are two different inflectional forms that belong to the same lexeme. It is often useful to distinguish word forms and lexemes by using italics and small capitals; so the lexeme book covers the forms book and books, and the lexeme write covers not only the form write but also writes, wrote, writing and written. Lexemes allow generalisations to ignore inflectional differences, as when we say, for example, that enjoy licenses an object (which is equally true of all its inflectional variants). Another important distinction contrasts word ‘types’ with word ‘tokens’. Types are stored items in the language, whereas tokens are examples of these types used on particular occasions or in particular texts. For example, The cat sat on the mat contains six tokens of five types, including two tokens of the type the. The ‘type-token’ ratio of a text is a familiar measure of the range of vocabulary in it.

word class. Words can be classified grammatically as nouns, verbs and so on. Like any other classification, these classes take account of a wide range of characteristics, and not just a single one which might be used as a simple criterion for recognising members. In the case of words, these characteristics include:

  • syntax: nouns and verbs combine with different kinds of words; for example, in English we use adjectives to modify nouns, but adverbs to modify verbs (recent accident, but happened recently).
  • morphology: nouns and verbs have different inflections: nouns inflect for number (singular vs. plural), while verbs inflect for tense (present vs. past).
  • meaning: for example, a noun may refer to a concrete object or a person, but this is not possible for a verb; in contrast, verbs tend to be used for events and states. However, it is important to stress that meaning, on its own, never provides a reliable basis for distinguishing one word class from another.

This idea that class membership involves many different criteria leads us to two important principles::

  • When you’re deciding how to classify a word, its meaning is often relevant and helpful, but it’s never the only thing to pay attention to: you always need to pay attention to its grammar — its syntax and morphology.
  • Nor is it enough to look at how the word is being used (syntactically) in a particular case. Instead, you have to consider its full range of possible functions. For instance, in big grammar book, the words big and grammar are both being used in the same way (to modify the noun book), but they actually belong to different word classes because they have different ranges of possible uses: big is an adjective because we can also say: This book is very big, but not: *I like big; but grammar is a noun because we can also say: I like grammar, but not: *This book is very grammar.

Using these multiple criteria, grammarians generally distinguish the following primary word classes for English:

Members of these word classes combine in sentences, but there are many other ‘minor’ classes of words which tend to be used in isolation, such as interjections (e.g. woops!), greetings (hello), fillers (oh) and filled pauses (erm).

Each of these main word classes may be further subdivided into secondary classes; for example, verbs can be classified as auxiliary verbs, among which we recognise the modal auxiliaries. Primary word classes were traditionally called ‘parts of speech’, but this rather opaque term is becoming much less common. The list of primary word classes presented in the glossary differs from the traditional list at three points:

  • The traditional adjective class is divided into two primary classes:
    • determiners (the traditional subclass of limiting adjectives)
    • adjectives (the traditional subclass of descriptive adjectives).
  • The traditional conjunction class is likewise divided into:
    • coordinators (the traditional coordinating conjunctions)
    • subordinators (the traditional subordinating conjunctions).
  • Conversely the traditional pronoun (a part of speech or primary word-class) is treated here as a secondary class of noun.

word family. The ‘words’ (technically, lexemes) in a word family are normally derived from a single ‘root word’; for example friend acts as the root word which holds together friendly, unfriendly, friendship, friendliness and befriend. As in this example, the lexemes in a word family may belong to different word classes and have quite different, though related, meanings, in contrast with the inflectional forms of a single lexeme.

word form. See word.

word order. One of the important parts of syntax is the order of constituents. The term word order is often used for something that could be better called ‘constituent order’. The order found in sentences is partially free, and partially covered by rules. For instance, English (an ‘SVO’ language) has relatively strict rules for the positions of a verb and its subject and object, and for the modifiers of a noun (adjectives before the head, preposition phrases after it); so I can’t stand self-important people with loud voices allows very little freedom in the order of words. However some freedom is possible even in English thanks to operations such as fronting, which, given a suitable context, would allow Self-important people with loud voices I can’t stand. Adverbials also tend to be free to move, as in Irene does it quite frequently – Irene quite frequently does it – Quite frequently Irene does it.

word token, word type. See word.

yes-no interrogative. See clause type.