secondary menu


This slideshow requires JavaScript.

The ‘Wanna’ Puzzle

In casual speech, English speakers commonly contract want to into wanna. This seems like it should be easy to explain: want+to=wanna, right? But in fact—as linguists often find when they study so-called ‘lazy’ or ‘sub-standard’ language—we actually use wanna in subtly complicated ways. Compare:

(a)  Contraction good:
Who do you wanna-contracts-whitebg dance with?
Which dancer do you wanna-contracts-whitebg see?
Who do you wanna-contracts-whitebg go to the games with?
(b) Contraction blocked by intervening word:
Who do you wanna contraction illustration dance with?
Which dancer do you wanna contraction illustration see?
Who do you wanna contraction illustration go to the games with?
(c)  Contraction blocked—but why?!
Which dancer do you wanna-contraction-blocked-no-gap-whiteBG dance next?
What team do you wanna-contraction-blocked-no-gap-whiteBG win the series?
Which party do you wanna-contraction-blocked-no-gap-whiteBG win the election?

Examples like (b) are easy to understand: obviously you can’t contract to wanna when another word gets in the way (want us to…). But examples like (c) are much more puzzling: no word intervenes—and yet, even the most laid-back speakers will not make want to into wanna in (c). Why not? What gives? What’s the difference between (a) and (c)?  This is the famous wanna puzzle. And if you ‘wanna’ know the solution, just read the rest of this post!

Background: how linguists see ‘gaps’

Every profession has its super-power. For people who study sentences (‘syntacticians’), their power is that they learn to see beyond the flat string of words, and behold a complex ‘web’ or ‘tree’ of subtle connections underlying each sentence. They see how words group into phrases; how the phrases form structures; and how the structures are constrained by complex principles. And one more thing: they also learn to see how certain words connect to invisible gaps.

That last point is especially important for understanding wh-questions. Ask a linguist to represent the ‘hidden structure’ of a wh-question, and for each wh-word they will mark a connection to a little linked ‘gap’, like this:

  • Whok did John leave with __k?
  • Whok did Mary annoy __k?
  • Whok do you think [__k is going to win the games at Sochi]?

The gaps are silent, of course. To figure out where to put them, linguists look for the ‘slot’ that the wh-word asks you to fill, when you answer:

  • Whok did John leave with __k? (answer: He left with Mary)
  • Whok did Mary annoy __k? (answer: She annoyed Sue)
  • Whok do you think [__k is going to win the games at Sochi]? (answer: I think Russia is going to win).

Why do linguists bother to highlight these connections? The reason is simple: that particular connection, the special meaning relation between a wh-word and the position it ‘asks about’, has an effect on a whole bunch of linguistic phenomenon—including, as we will now see, on wanna contraction.

The Solution: Gap blocks Wanna

Look again at the wanna-puzzle data—but this time, with those ‘linked gaps’ highlighted:

(a) Contraction good:  nothing intervenes
Whok do you wanna-contracts-whitebg dance with __k?
Which dancerk do you wanna-contracts-whitebg see __k?
Whok do you wanna-contracts-whitebg go to the games with __k?
(b) Contraction blocked by intervening word:
Whok do you wanna contraction illustration dance with __k?
Which dancerk do you wanna contraction illustration see __k?
Whok do you wanna contraction illustration go to the games with __k?
(c) Contraction blocked by the intervening gap
Which dancerk do you wanna contraction illustration dance next?
Which teamk do you wanna contraction illustrationwin the series?
Which partyk do you wanna contraction illustration win the election?

As you can see, a simple generalization now emerges—and it applies, systematically, to all cases of wanna contraction:

  • If nothing intervenes, want and to can contract into wanna.
  • If anything intervenes (even if it’s a silent-but-deadly ‘wh-linked gap’) then this blocks wanna contraction.

There are some theory-internal complications, but that, in a nutshell, is the solution to the wanna puzzle. You just have to ‘mind the gap’!


  • There may actually be several other types of ‘gaps’ besides the wh-linked gaps discussed here. Linguists call them all ’empty categories’. To distinguish the ones discussed above, linguists call those ones ‘case-marked‘ empty categories.


  • The other types of empty categories do not seem to block wanna contraction. Because of this, the technical constraint is like this: a ‘case-marked’ empty category blocks ‘wanna’.


  • Wanna contraction is also blocked in sentences like this:
    • Do you want [ [to be in love] to be a crime] ?

    This is probably unrelated to the main wanna puzzle; what’s happening here is that the to, though adjacent in word order, is too far ’embedded’ in the structure of the sentence.


  • The wh-linked gaps discussed here do not block ALL forms of contraction. Some other words can contract fine, even when a case-marked empty category ‘intervenes’. For example,:
    • Whok do you think [__k‘ll win the race]?

    It’s not clear why the different types of contraction behave differently. A linguist’s work is never done!



As far as Dr. Dexterous can tell, the first published discussion of the wanna puzzle is by George Lakoff, in this paper:

  • Lakoff, George (1970). Global Rules. Language 46: 627–639.

The idea that a wh-linked ‘gap’ (technically called a ‘trace’ or ’empty category’) blocks wanna contraction is a simplified version of a proposal by Noam Chomsky, in this paper:

  • Chomsky, Noam (1976). Conditions on Rules of Grammar. Linguistic Analysis 2: 303–351.

The fact that other types of empty categories can intervene, and the full technical statement of the constraint, appear in a paper by Osvaldo Jaeggli:

  • Jaeggli, Osvaldo (1980). Remarks on To Contraction. Linguistic Inquiry 11: 239–246.

Additional structural constraints on wanna contraction are noted, and a big debate about how to handle the whole thing is raised, here:

  • Postal, Paul M. and Geoffrey Pullum (1982). The Contraction Debate. Linguistic Inquiry 13: 122–138.

And a nice summary of it all, including another alternative (though somewhat technical) solution, is here:

  • Goodall, Grant “Contraction”, in The Blackwell Companion to Syntax, Volume I Edited by Martin Everaert, Henk van Riemsdijk Copyright © 2006 by Blackwell Publishing Ltd

One Response to wanna

  1. wondering December 11, 2014 at 7:57 pm #

    Is “gonna” another example of the same puzzle, or is it different somehow?

    Are you gonna go to the game?
    Which movie are you gonna see?
    Are you gonna be here when I get back?

Leave a Reply

HTML Snippets Powered By :