How do contractions work in syntactic movement?

Specifically when the surface structure uncontracted would be ungrammatical. Eg: "Don't turtles live forever?" (Do turtles not live forever/*Do not turtles live forever).

EDIT - Sorry if it's broad and vague, perhaps I simply have a poor grasp of x-bar theory in general ^_^' What I mean is, what's going on with the negation in a sentence like this? We know that it can't be put before the subject as in *Do not turtles live forever, and we know that when we say "Don't turtles live forever?" we really mean "Do turtles not live forever?" so why is the contraction "Don't" acceptable? If the surface structure was "Do turtles live forever" it would seem as simple as moving the dummy do into the CP head position, but the contraction/negation thing throws me off.

• This is a very broad/vague question. What exactly do you mean by "work in syntactic movement"? Sep 18, 2017 at 8:16
• Does the edit clarify? Sep 18, 2017 at 9:50
• A 1983 Zwicky and Pullum paper arguing that "n't" is an inflectional suffix: web.stanford.edu/~zwicky/ZPCliticsInfl.pdf Sep 18, 2017 at 13:35
• @Ellie Yes, much better :) Sep 18, 2017 at 14:25

Summarizing the paper by Zwicky and Pullum commented by @sumelic above: They suggest that most contractions are clitics, but <-n't> is an inflection.

Most English contractions, such as <-'s> <-'ve> <-'d>, behave like clitics that can be directly substituted for the words they abbreviate. One of their examples is "The ball you hit's just broken my dining room window" where the <-s> for has is cliticized to 'hit'. One can even string multiple clitics, like "I'd've done it if you asked me."

However, the contraction <-n't> for not does not pattern like a clitic, but rather behaves an inflectional suffix; specifically as an inflectional suffix on finite auxiliary verbs. This supports your own example, where "Don't turtles live forever?" cannot be expanded to "*Do not turtles live forever?", and also precludes things like "*I'dn't".

I'm not convinced the notion "clitic" is really needful to explain what is going on. Some syntactic rules depend on what the words are, and you can't always trust traditional English orthography to tell you the truth about where the words are. Why should should you? Probably printers made up those rules.

The rule of subject auxiliary inversion moves a word to before the subject. In your example, "don't" is a word, but "do not" isn't a word, and that's why we do not get *"Do not turtles live forever?" Two words moved, not just one.