In sentences (1-3), the verbs in italics are usually called 'auxiliary verbs'.
(1) It is raining.
(2) You have been overcharged.
(3) It must have been raining.
These auxiliary verbs are distinguished from other verbs by two characteristics:
(4) It is raining. - Is it raining?
(5) She wrote a book - BUT NOT: *Wrote she a book?
The trouble is that these two characteristics are separate and each define a different group of verbs, to which grammarians give distinct names:
(6) She was/got chosen for the job.
(7) She was/kept talking.
(8) She ought/started to talk.
(9) She will/helped wash up.
(10) She is ready. - Is she ready?
(11) She has some money. - Has she any money?
The clear cases of auxiliary verbs combine the two sets of properties, so we can provisionally define auxiliary verbs like this:
The auxiliary verbs defined by this criterion are:
The characteristics of operators apply to English, but they are irrelevant to other languages. However they are fundamental to English grammar and cannot be ignored. The special role of operators has evolved in English since the time of Shakespeare, but it is now firmly established in all varieties of modern English.
(12) He was run over by a bus. - Was he run over by a bus?
(13) He got run over by a bus. - BUT NOT: *Got he run over by a bus?
(14) It is not raining. - It isn't raining.
(15) He is not ready. - He isn't ready.
(16) BUT NOT: *He got not run over by a bus. - NOR: *He gotn't run over by a bus.
(17) He did not get run over by a bus. - Did he get run over by a bus?
(18) BUT NOT: *He did not be run over by a bus. NOR: *Did he be run over by a bus?
(19) NOR: *He did not be ready. NOR: *Did he be ready?
(20) People never get run over. BUT NOT: *People get never run over.
(21) People never are run over. - People are never run over.
(22) He never is ready. - He is never ready.
(23) It is raining. - It's raining.
(24) It keeps raining. BUT NOT: *Itk'ps raining.
(25) He is ready. - He's ready.
(26) He will (NOT: *wills) come.
We have defined auxiliary verbs as operators which are also catenatives, which has the effect of excluding operators that are not catenatives - the uses of be and have in (10) and (11). This has the advantage of preserving the traditional link between 'auxiliary verb' and the support role, but it faces a serious problem: the list of words picked out as auxiliaries is exactly the same as the list of operators! This is because be and have both have catenative uses as well as their non-catenative ones. Grammarians adopt two different positions on this problem:
On balance the second of these approaches seems preferable, but if we do adopt it, which of the terms should we preserve?
Neither of these options is attractive, but there is a way forward. Suppose we distinguish between a general definition of auxiliaries (which applies across languages) and a language-particular definition of the auxiliaries in some particular language (which will obviously have to match the general definition). The general definition builds on the 'supporting role' of catenatives, but also says that there must be some other grammatical peculiarity to distinguish auxiliaries from other catenatives:
An auxiliary verb is a verb that combines two characteristics:
An auxiliary verb is a verb that combines two characteristics:
Notice how the definition for English is a slight modification of the general definition - it changes supports to can support, and it supplies a particular range of other distinctions. The support role is still relevant, because every auxiliary verb can be used to support another; but it is not crucial because the main distinctive load is carried by the other distinctions. The consequence of this definition is that a sentence like He is ready. contains an ordinary and straightforward auxiliary verb as far as the English definition is concerned, but this auxiliary verb is untypical in comparison with the general definition.
The crucial question is how to analyse sentences like (10) and (11), where be and have have non-catenative uses. There are two possible answers:
Unfortunately sentences like (10) and (11) are extremely common, so it is hard to avoid this choice; and it is even harder to fudge it by finding a compromise position in between!
Other languages also have verb classes which fit the General Definition, with 'other distinctions' that vary from language to language.
(27) Paul les a mangés. 'Paul has eaten them.'
(Note that les, 'them', is attached to a, 'has', but belongs as object to mangés, 'eaten'; this is not allowed for other auxiliary-like verbs such as vouloir, 'want': Paul veut les manger, not *Paul les veut manger, 'Paul wants to eat them.')
(28) Nos vienen a ver. 'They are coming to see us' (literally, 'Us they-come to see.')
(29) Ich weiss, dass Paul gekommen ist. 'I know that Paul come is' - i.e. 'I know that Paul has come.'
(30) Ich weiss, dass Paul zu verstehen schien. 'I know that Paul to understand seemed' i.e. 'I know that Paul seemed to understand.'
(31) Ich weiss, dass Paul hofft, zu verstehen. 'I know that Paul hopes to understand.'
And so on through other languages, each containing a small sub-class of verbs which can support other verbs and which follow different rules of grammar from all other such verbs.
The main point of this discussion is to remember that each language
sets its own defining characteristics for 'auxiliary verbs', although there
are some similarities across languages. In the case of English, this means
that auxiliary verbs are defined by characteristics such as their ability
to form negative and interrogative sentences rather than by their ability
to support another verb.
Nearly all these references opt for the first solution, in which 'operator' is used for the special characteristics and 'auxiliary verb' is reserved for operators which also support other verbs. The exception is Hudson (1998), which argues for the second solution.
Greenbaum, Sidney. The Oxford English Grammar. Oxford University Press, 1996, 153-6, 248-51.
Huddleston, Rodney. Introduction to the Grammar of English. Cambridge University Press, 1984, 136-143 - especially 140-3.
Huddleston, Rodney. English Grammar: An Outline. Cambridge University Press, 1988, 45-7.
Hudson, Richard. English Grammar. Routledge, 1998, 51-60.
Quirk, Randolph; Greenbaum, Sidney; Leech, Geoffrey and Svartvik, Jan. A Comprehensive Grammar of the English Language. Longman, 1985, 64, 79-82, 120-148.