When I draw the syntactic tree of a sentence with "not", what kind of component would the "not" be?


Jane did not go to school

  • It's an adverb; an absolute negation marker. In a negative clause, it modifies the verb. Your example contains 'verbal negation' which is marked by modification of the verb "did".
    – BillJ
    Commented Jul 1, 2016 at 16:59
  • 4
    I think the question is on-topic. It asks about the linguistic explanation of a general syntactic phenomenon; also I don't think it's a classical "please draw me a syntax tree" request, because the queston of the syntactic analysis of negation is a rather general theoretical question and the OP is probably not just interested in one readily copy-pasteable tree, but rather how negation could be accounted for at all. @Dotan Reis Maybe you could still provide an attempt showing what your suggestion would be, so we can meet your question more specifically. Commented Jul 3, 2016 at 11:52
  • I was just wondering where this fits in a syntax theory - should it be a specifier for the verb (Jane [not go] to school) or to the whole sentence (not [Jane went to school]) etc
    – Dotan
    Commented Jul 3, 2016 at 15:12
  • 2
    I think that @jlawlers comment on eijen's answer clarifies the discrepancy between logical and syntactic placement of the negation pretty well. Also, be aware of the difference between a specifier (which is the third phrasal component besides the head and the complement in X-bar theory) and modification (which is a syntactic function, usually being realised by adjunction). Commented Jul 3, 2016 at 16:58

3 Answers 3


Totally depends on your syntax theory.

Some prefer to do it with a NegP, as suggested bei @eijen:
enter image description here

Others assume the negation to be in I:
enter image description here

And then again you could see negation as an adverb modifying the VP, as suggested by @BillJ:
enter preformatted text here

There really isn't a uniform answer, because that depends so much on your syntax theory (already whether you assume that there is something like TP and little v, go with I instead, maybe something even different, ...), so you'd need to specify what framework you are in.

  • Note that if the sentence used 'can' instead of 'did', the 'not' would have to modify 'can' rather the verb (or lead to two possible meanings). This might point to the better model. Since 'not' takes only one 'bit' (inhibition rather than excitation), it may not be stored like other 'adverbials'.
    – amI
    Commented Jul 1, 2016 at 19:50
  • @aml So which tree would be the better model in your opinion? I'm not sure I fully understood what you mean by your first sentence; I guess it is the ambiguity in the surface form "She can not go" between the logical forms ¬can(go) and can(¬go)? This is an idiosyncractic property of English syntax and probably simply can not be universally resolved in a "better" syntactic tree because there are always two readings possible, which however should not disturb the theory's general functionality - I recently had a similar discussion here. Commented Jul 2, 2016 at 15:25
  • It's really better if we don't encourage off-topic questions by answering them like this...
    – curiousdannii
    Commented Jul 3, 2016 at 11:21
  • @curoiusdannii Why do you consider the question off-topic? It is clearly about the linguistic explanation of a general syntactic phenomenon - the only thing I would criticise about the question is the OP not providing a proposal on his own, but possibly he/she just doesn't have any idea (I don't know how much background the OP has in syntax theory), and if others then get interested in discussing the topic more detailled, why not. Commented Jul 3, 2016 at 11:23
  • 1
    Syntax trees questions are off-topic precisely because of what you wrote in your first sentence: "Totally depends on your syntax theory". There are as many ways to draw as syntax trees as there are linguists. Questions need to be so much more specific than this one is to be on-topic.
    – curiousdannii
    Commented Jul 3, 2016 at 12:38

Logically, "not" is a sentential adverb. Grammatically, in English, it is an auxiliary verb suffix. The disparity between its logic and its grammar explains why it is so difficult to classify, especially for those grammarians who cannot distinguish logic from grammar.

A grammatical modifier is an element which when added to something of a certain grammatical category gives a new element of that same category. In logic, negation and the modal operators are treated as being added to sentences (or, that is, propositions) to create new sentences. If p is a sentence, then ~p is also a sentence. Although logicians don't use this terminology (they generally call ~, Nec, etc. "operators"), this makes negation a grammatical modifier. Grammarians generally call modifiers of nouns "adjectives" and other modifiers "adverbs", so it is reasonable (though vague) to call "not" an adverb, judging from its logic, since it modifies something other than a noun (viz., a sentence).

In English grammar, though negative "never" can reasonably be classified as sentential adverb, comparable to "possibly" or "maybe", "not" itself is not at all a grammatical adverb. Instead, it is suffixed to a finite auxiliary verb like "is", "can", "does", to create a new finite auxiliary verb ("isn't", "cannot", "doesn't", and so on).


not is the head of a NegP in English. See ch. 6 of Beatrice Santorini's syntax textbook for the history and reasons for why not is understood as a NegP in English.

  • It's a marker. In a negative clause, it modifies the verb. You could call it 'head' of an AdvP if you liked,
    – BillJ
    Commented Jul 1, 2016 at 17:04
  • Muntañá, 2008: In addition, not and the contracted form -n’t, which are two possible phonological realisations of the sentential negative marker are assumed to be contextual allomorphs of the same syntactic head." (emphasis mine) It's marker in a syntactic head.
    – eijen
    Commented Jul 1, 2016 at 17:09
  • Then why does Do-support happen? There's tense-lowering in English, so something must be blocking the movement of the T.
    – eijen
    Commented Jul 1, 2016 at 17:17
  • Do-support is required in negative clauses with a lexical verb. Verbal negation is enabled by the modification of dummy "do" by the negator "not".
    – BillJ
    Commented Jul 1, 2016 at 17:31
  • 1
    Logically, it's the outer operator in ¬ Go-to-school (Jane), which is to say it modifies (or operates over, if you prefer) the entire clause. The placement of the negative in the sentence is a consequence of syntax, not modification.
    – jlawler
    Commented Jul 1, 2016 at 18:04

Not the answer you're looking for? Browse other questions tagged or ask your own question.