Sticky magnetic negativiser

We already discussed the basic building blocks of Syntax, for all languages – which boils down to a structure known as an XP (X being a variable and P standing for Phrase, so XP is a ‘phrase of any kind’). An XP – i.e. every Phrase (a Phrase in the linguistic sense, not the grammatical sense, thought the two are related), has the following structure. It has a Spec, a position into which another phrase can move; a Head, where its ‘active’ element may reside; and a Complement, where another Phrase may (or may not) naturally occur. The Syntax of a clause is a large tree of XPs, some of whose elements may have moved to different Spec positions.

Simplified XP

We’ve also already seen the underlying components of sentence structure: The CP (Complementiser Phrase, where clauses and questions happen), The IP (Inflect Phrase, where the Subject lives), and had a glance at Adverbs and the Verb Phrase. However, one of the most interesting and puzzling elements in English Syntax is what is called the Negativiser (or NEG). The most common Negativiser of all is “not.

sentence with neg2

Neg occurs immediately after the IP; it is considered to have its own phrase (or XP), of course, as it must – every constituent of the Clause has its own Phrase. Unlike Verbs (but like Adverbs as we will later see), the content of the Neg is considered to reside on the Spec of its XP (where phrases go), not the Head. This is unusual, since the Neg is a very ‘active’ element, and ‘active’ things typically happen in the Head. This is not the only unusual thing about the Neg – it has a number of oddities.

Unlike other phrases, the Neg very seldom moves independently. Other phrases (nouns and adverbs) often move freely to other Spec positions (we saw some of this in the post on Sentence Structure); the Neg strongly prefers to attach to another element – it is ‘magnetic’. It’s almost unique in this respect.

As you see in the first diagram below, the natural position for the Neg (“not,” in this case), is after the Modal or first Auxiliary (example (a)), which gives the most Syntactically natural negative sentence, an example of which is “I would not normally call you.” As noted, the Neg loves to attach to other sentence constituents – most obviously, whatever is in the Head of IP. If there is something in the Head of IP, the Neg will often (almost always in spoken English) encliticise* to it (example (b)), giving us, “I wouldn’t normally call you.”

(* ‘Clitics’ – of which ‘encliticise’ is the verb, are those bits of words that attach to another, such as ‘ve, ‘d, ‘ll and ‘nt).

Neg also has a strong tendency to attach to pronouns and adverbs, so strong in fact that it sounds unusual to the point of being ungrammatical when this doesn’t occur. Example (c) in the diagram below is a highly unusual sentence in English. It’s not grammatically incorrect, but preventing the Neg from combining with the pronoun, someone, causes  a lot of tension and is not a natural sentence. If the Neg is allowed to combine to form the pronoun “no one”, the tension is resolved.

neg sentences 1

The same is true for examples (d) and (e) in this diagram. Not combining with the adverb (ever, in this case), causes a great deal of tension, but combining resolves the tension.

Neg prefers pronouns to adverbs, for example:
      No one will ever call you,
sounds much more natural than:
      Someone will never call you.

There are several very unusual things about these examples. Firstly, tension in Syntax is usually caused by movement; and the most Syntactically natural, formal, least tense arrangement is the one with little or no movement. Neg does the opposite – tension is resolved by movement. Furthermore, Chomsky has said that all movement is leftward movement – not everyone agrees entirely with this, but Head-to-Head and Spec movement are definitely always to the left (there’s an obvious reason for this which I may discuss in a later post) – however, the Neg doesn’t always move left; it seems to move randomly to wherever it can attach to something.

And this raises the most interesting point, that what Neg is doing is both Syntactic (to do with the sentence structure) and Morphological (to do with word formation). In other words, it is Morpho-Syntactic and these are Morphosyntactic processes. Syntactic processes and Morphological are each often observed, but processes that are both Morphological and Syntactic are not very commonly observed in English, which makes this is very interesting.

Here’s another example.

neg sentence object4

Once more, despite apparently being the most Syntactically natural arrangement, example (a) above (“I have not told someone”), contains a certain amount of tension – to the point of almost being grammatically incorrect. To resolve the tension, the Neg can do two things here. It can combine with the Object pronoun (someone) to form no one; or it can stay in its natural position and exercise a Morphological effect on the pronoun someone to transform it into anyone, in a ‘not… anyone’ construction.

One thing my Syntax professor stressed is that ‘downwards’ movement (rightwards movement in the above example, bearing in mind that the above are simplified representations of Syntax trees that progress downwards) is not possible (“no! Constituents can’t move down!” she would cry, if I made a mistake in my Syntax diagrams). At least in pure Syntactic terms, all movement is leftwards movement (you have to trust this, it has been proven). It isn’t possible, Syntactically, for the Neg to enter the Verb Phrase – the English Verb Phrase (as we will see in a later post) is a fortress – constituents can leave, but nothing can enter. So this is an extremely interesting process.

One can even think up a possible (albeit unusual) example like the following:

neg sentence object not a soul

“I have told not a soul” is an unusual sentence in English. It isn’t natural, but it’s not completely ungrammatical. What is happening here looks for all the world like a Syntactic process – Neg (“not”) moving not just into the Verb Phrase, but into a Noun Phrase within the Verb Phrase. This is completely impossible according to the accepted rules of Syntax. Unlike ever becoming never, it doesn’t seem to be a Morphological process (although perhaps some very complex Morphosyntactic process is happening). It’s quite a mystery; and very interesting.

Finally, here’s an interesting example of Neg combining with the Complementiser that, and exercising its Morphological magical over the Object pronoun, to form a “not… anyone” clause construction: “…not that I have ever told anyone.” This is an idiomatic construction, which has a slightly different meaning from a simple negation of the that-clause, but it’s still somewhat similar in meaning

For example, one could say:
      “you and I have dark secret, that I have not ever told someone,”
or one could emphasise that the secret hadn’t been told by saying:
      “you and I have a dark secret others would seek to discover, not that I have ever told anyone.”

The second of these is an interesting construction that is a pleasure to contemplate in light of the above. It’s a charming, playful expression in English and from a linguistic perspective has both a Syntactic and Morphological component: the Neg is jumping all the way to the Spec of CP to join with the Complementiser “that,” much as it moves to combine with pronouns and adverbs; and the “not anyone” construction is occurring via a change to the Object pronoun in the Verb Phrase.

not that i ever told anyone2

In fact, it may be that there is no Syntactic process happening here at all; that Negativisation takes place at the Semantic, pre-verbal level; and that the Neg can naturally appear at various points in the Syntax depending on the Semantics of the original thought.

Note that it is not unusual to have an MWE (a Multi-Word Expression) in a Spec position, and it causes no tension, unlike the MWE in the Head position in the earlier example. The Head position is active and very particular about what sort of content and what sort of movement is allowed, whereas Spec is a very general, flexible location for any Phrase (XP) to move to.

Questions with a negativiser

I looked at the Syntax of questions in the post on sentence and Clause structures. Questions with negativisers work much as one would expect them to. In example (a) below, just like regular questions, there is Head-to-Head movement of the Modal Verb from the Head of the Inflect Phrase to the Head of the Complementiser Phrase, leaving the Neg in its natural position. Once more we can see that standard formal English corresponds to the most natural, tension-free ordering of English Syntax (example (a):

Would someone not have seen something?

In spoken English, of course (example b), the Neg typically encliticises to the Modal (or whatever Verb is occupying the Head of IP), the Negativised Verb then moves to the CP.

Example (c) is interesting. This is an example of ‘hyper correct’ or ‘over correct’ formal English. English speakers make mistakes such as this when they are trying too hard to sound formal or educated. There are two words trying to occupy one Head position in the CP; and this construction sounds far from natural. However, as awkward and unnatural as it sounds, it doesn’t actually sound completely ungrammatical, which is interesting, in that a single ‘word’ in Linguistics (a ‘Lexical Item’ as they’re called) doesn’t need to correspond to a single word as commonly understood (an ‘Orthographic Word’), although, language being a natural phenomenon, they usually do correspond. In this case, the brain allows that “would not” can form a single Head, although not comfortably.

neg questions

Examples (d) and (e) again show the same interesting Morphosyntactic properties of the Negativiser, the Neg combining with either the Modal or inflected Auxiliary Verb (as a clitic), which then move together in Head-to-Head movement.

It’s noteworthy that, unlike other sentence components, the Neg always combines with another constituent when it moves. It never seems to move on its own. Even when you find “not” in a different location in the sentence, it is actually sharing a Spec or Head with another constituent. Other phrase types don’t do this.

Negative inversions

The following – a particular type of inversion (inversion is the name given in grammar to a sentence where the subject and verb change relative positions) – is to me one of the most fun and interesting things about Negativisers in English, and highlights their sticky magnetic quality. As noted, Neg is drawn to combine with other constituents, notably Pronouns and Adverbs and also Nouns, but as well as this, it seems to be drawn to the start of the Clause.

There’s a process called Topicalisation (which I may discuss in more detail later), in which elements of the sentence move into the Spec of CP (the start of the sentence) and become emphasised. This can occur with all kinds of elements (some simple examples were discussed in the post on sentence structure), among which are Adverbs and Prepositional Phrases. For example:

      Always I tried to be a good father.
      In a million years, you won’t see such a sight.
      There is no inversion in either of these examples.

However, when such adverbial elements (Prepositional Phrases are not adverbs, but they are adverbial – they modify verbs –  so both Adverbs and Prepositional Phrases are adverbial), combine with a Neg – as shown in the two examples below, and then Topicalise (move to the Spec of the Complementiser Phrase, the first possible position in the Clause), it seems as if – as they pass the Head of Inflect Phrase, where the Modal or the Inflected Auxiliary Verb is located – the negativised adverbial pulls the Verb out of its position and forces it to move to the Head of the CP. The diagram below shows what happens. This to me is an interesting and exciting example of the sticky magnetic power of the Negativiser that it pulls the Verb out of position as it passes, although with a non-negative expression, the Verb stays where it is.

For example:
      I have never seen such a thing.
      Never have I seen such a thing.

Both movements are necessary. *”Never I have seen such a thing,” is not grammatical in English. The diagram below depicts this process for an Adverb and a Prepositional Phrase.

neg inversion

You can see how the negativised elements seem to pick up the Verb as they pass.

Negatives and double negatives

Double negatives are ungrammatical in English, inasmuch as they can be used in very specific circumstances, but in ordinary speech they are ambiguous and confusing. For example:

      I wouldn’t never do that,
is simply ungrammatical.
      I didn’t do nothing,
can mean something specific and emphatic (‘the claim that I did nothing is false’), but is not grammatical in normal speech.

There are some adverbs that may not appear to be negative, but can be shown to be so by the fact that they trigger the feeling of a double negative when used with a Negativiser. For example ‘little’ (as as adverb); and ‘hardly’.

      I didn’t do little,
sounds like a double negative; as does,
      I hardly hadn’t finished.

Unsurprisingly, then, both of these adverbs also trigger inversions when Topicalised. So one can say:

      He knew little,
but when Topicalised one must say:
      Little did he know, that…
      She had hardly finished, when…
      Hardly had she finished, when…

Interestingly, perhaps due to their emphatic quality, but both of these adverbs must form part of a Subordinate Clause, to be juxtaposed with another statement. They do not work as statements on their own. For example, hardly had she finished when the phone rang,” or little did he know, but his lover had made other arrangements.”

Note that this a Syntactic, not a logical problem with double negatives. Double negatives sound perfectly grammatical if you split them across two clauses, for example in a Cleft construction.

For example:
      I didn’t do nothing,
sounds bad, even if it is logical in the context, whereas:
      It wasn’t that I did nothing,
sounds perfectly grammatical. As a Cleft construction – one statement split into two clauses – English Syntax comfortably allows one Neg per clause.

      [ CLAUSE 1 It wasn’t  [ CLAUSE 2 that I did nothing ] ] 


Obviously, Negativisers in English are  not literally magnetic, they don’t bounce around the sentence grabbing hold of Pronouns and Adverbs and pulling Verbs out of position; but the movements and transformations that occur around Negativisers make it seem as if this is happening. There is still a great deal to be understood about the cognitive processes behind the transformations that take place in Syntax and Morphology; the ‘sticky magnetic not’ is just an amusing analogy.

Indeed, much of what happens with Neg may in fact be Semantic. The negativised sentence or clause may occur as a thought, in a pre-verbal state, before it is given form and structure in the Syntax; and the Neg may simply be occurring in one of a variety of Syntactically natural positions; and thus the process is essentially Morphological. However, the tension that occurs with different constructions is far more typical of Syntactic processes. Whatever is happening, these are just some of the beautiful riddles that can be observed in Linguistics.

(And if you find “not” interesting, you should read about “ne… pas,” the wild and crazy French Negativiser.)


Leave a Reply

Your email address will not be published. Required fields are marked *