Polarity

Clausal negation

The usual strategy for rendering a positive clause negative in English is to insert the negator (the word not) between the auxiliary and the main verb. If no auxiliary was present in the basic clause, the dummy do is used. For example, Mel has not taken units in creative writing or cake decorating. Mel does not plan to take such units in the future.

Clauses that are identified as having negative polarity can be divided into two types. The distinction relates to the place in the clause where the negation occurs.

Verbal negation

Clauses are negated using the form not or the contraction n't. In clauses with a primary verb, the negator follows an auxiliary. If the basic clause does not contain an auxiliary, the 'dummy verb' do is inserted into the clause. The only verb that does not require the dummy do when it occurs without an auxiliary is be.

In negative imperative clauses the dummy do is always required:

In subordinate clauses the negator occurs without an auxiliary:

Non-verbal negation

It is also possible to mark clause level non-verbal negation using other negators besides not. There are two types of non-verbal clause level negators.

• Absolute negators include: no, none, nobody, no one, nothing, nowhere

• Approximate negators include: few, little, rarely, seldom, barely, hardly, scarcely

Consider the following examples:

Subclausal negation

There are many negation strategies that apply to particular words within the clause but do not render the clause itself negative. In clauses of this type the negation is said to be 'subclausal'. Negative derivational prefixes include dis-, non-, in-, un-, and suffixes such as -less allow us to form negative words that are associated with subclausal negation.

It can be helpful to use the tests mentioned below to determine whether or not a clause is negative because sometimes the same forms can negate either the whole clause or an element within it. Consider the following examples:

Distinguishing negative and positive clauses

Clausal polarity can be tested by using a tag question; the tag will usually have the opposite polarity of the clause; for example, the grammatical tag question for Fortunately for us Gina doesn't like celery is does she? (i.e. positive tag, negative clause), while the tag for Unfortunately for us Gina likes celery is doesn't she? (i.e. negative tag, positive clause). Positive clauses can also often take positive tags, for example, Gina likes celery, does she?, but negative clauses can ONLY take positive tags (for example, *Gina doesn't like celery, doesn't she? is not grammatical).

Another easy test for identifying negative clauses in English is to add a constituent introduced by not even and see if it makes sense. Only negative clauses can take this type of constituent, as the following examples show:

A further special characteristic of negative clauses is that they can be linked by neither or by nor and cannot be linked by so.

The dreaded double negative (I don't mean nothing by it...)

A prescriptive grammar of standard English may insist that sentences such as the following are unacceptable. (They're probably going to look a bit funny written in black and white – try them in a regional accent of your choice – they're common place in a whole range of varieties of non-standard English.)

Often the explanation supplied in these texts is that the negators cancel each other out:

Mostly however, it is clear to English speakers that the negation stands. There are many languages (including French, Russian and Polish among others) where multiple negation is expected and in these languages we can talk about 'negative concord'. Non-standard Englishes allow negative concord as a means of reinforcing the polarity of the clause.

And in fact, there are several contexts in which a double negative is acceptable in standard English – and only in some cases is the resultant meaning positive: