Most English speakers, I think, would agree that these two sentences are essentially the same:
- I want to go to the party! \(\Leftrightarrow\) I wanna go to the party!
Now, I need to make a confession: I don’t know a single thing about phonetics or phonology. But, in my ignorance, I wouldn’t have thought twice about (1) — I would have just said something like:
This is some basic phonetic thing! The more frequently you say something, the more you can fudge the pronunciation. No big deal.
That is… until last month. You see, there are times when “wanna” is flat-out wrong.
- Who do you want to kiss that puppy? (I want Douglas to kiss that puppy!)
- Who do you wanna kiss that puppy?
Would I understand someone who said it? Of course. But, if I were speaking naturally, I just wouldn’t say it. Try saying both the sentences in (1) slowly; now try (3) — it just sounds strange.
And it’s not simply that some sentences can’t use “wanna” — the meaning of the sentence seems to matter! For example, let’s say you and your friends are going to watch a very exciting basketball game.
- Let’s watch the game! Who do you want to win? (I want the 76ers to win!)
- Let’s watch the game! Who do you wanna win?
But now, picture some very shady team owners, gambling their players away.
- Who do you want to win? (I want to win Joel Embiid!)
- Who do you wanna win? (I wanna win Joel Embiid!)
Clearly just because “want” and “to” are near each other doesn’t mean you can use “wanna”! This isn’t a phonological property of English; there are grammar rules dictating when and where “wanna” can pop up in a sentence! The question is, what are those rules?
Today’s post will try and tackle that question in three acts.
| Act 1 | We look at a widely repeated, very popular, old-school answer. |
| Act 2 | We present a list of (very alarming) sentences that the old-school answer doesn’t explain. |
| Act 3 | Finally, we arrive at a more satisfying answer to the wanna-question. |
The Old-School Answer
Act 1 — The explanation of “wanna” you are sure to find in your textbook.
So-called wanna-contraction (or to-contraction) has been something of a hot topic ever since the 1970’s. At that time, some certain linguists at a certain university in Massachusetts were developing pretty groundbreaking theories about language.1 A prominent piece of those theories is something called movement.
- Movement
-
In a syntax tree, certain rules may cause nodes to jump to
different positions in the tree.
e.g. In English, questions cause verb movement:
“You can do it!” \(\Rightarrow\) “Can you do it?”
Let’s look at a very well-known type of movement: wh-movement.
Example: Consider the sentence “you do like him”. What if we didn’t know who was liked?
“You do like him”2 \(\Rightarrow\) “Who(m) do you like?”
| You do like him. | The auxiliary verb is “do”. |
| Do you \(\square\) like him? | In English, questions move the auxiliary verb to the first position. |
| Do you \(\square\) like whom? | For our question we want to use “who(m)” instead of “him”. |
| Whom do you \(\square\) like \(\square\)? | In English, “wh”-phrases get moved to the beginning of the sentence (even before the auxiliary verb). |
Now what does this have to do with wanna-contraction?
Silent Words
I’m not sure who first noted the idiosyncrasies with “wanna” usage, but Chomsky and Lasnik (1978) were the first to provide a captivating explanation. You see, in the theory of movement, words don’t just move — they leave something behind!
- Trace
-
A trace is a hypothetical silent word, which is left
behind whenever movement occurs in a tree. Traces are usually
marked with a “\(t\)”.
e.g. In “Who do you \(t_{\text{do}}\) like \(t_{\text{who}}\)?”, we see two traces, each corresponding to a different instance of movement.
Chomsky and Lasnik were looking for evidence that silent words3… you know… actually existed. While doing so, they saw these problems with “wanna”, and made a pretty interesting observation:
Whenever a trace is found between “want” and “to”, it seems to “block” contraction.
| I want to kiss him. Who do I want to kiss \(t_\text{who}\)? Who do I wanna kiss? |
There is no trace between “want” and “to”. |
| I want him to kiss the dog. Who do I want \(t_\text{who}\) to kiss the dog? Who do I wanna kiss the dog? |
There is a trace sitting between “want” and “to”, which blocks the contraction! |
| I want him to win Who do I want \(t_{\text{who}}\) to win? Who do I wanna win? |
Notice we have “want him to” in the first sentence. That “him” becomes a trace in the question form. |
| I want to win him Who do I want to win \(t_\text{who}\)? Who do I wanna win? |
If “wanna” would work in the non-question sentence, it will work in the question form as well. |
That’s what I’m calling the “old-school answer”: silent words are blocking “want” and “to” from contracting!
Problematic Data
Act 2 — A whole list of sentences the old-school answer doesn’t seem to address.
The old-school answer (that traces are the reason “wanna” isn’t always allowed) can be found in basically any textbook on syntax (Falk 2011); I, for one, learned about it from Carnie (2013, chaps. 5, p.369). And that’s frankly pretty odd, because it seems to be incorrect.
I don’t mean movement is dubious (that’s a whole other debate). It just doesn’t explain what’s going on with “wanna”. You see, there’s a whole host of other situations where “wanna” just doesn’t seem to fit.4
Coordination
When “gluing” verbs together with “and” or “or” (known as
coordination), we find that wanna-contraction fails.
- I don’t expect, need, or wanna dance.
- I don’t wanna or expect to dance.
It’s not clear why there would be a silent word blocking this contraction.
Noun-Form
There is a noun form of “want”, such as in “my only want is to be
remembered”. Even when to immediately follows this “want” we don’t
see any contraction.
- We cannot expect that want to be satisfied.
- We cannot expect that wanna be satisfied.
I find (11) almost laughable. It’s clear that “want” and “to” simply being next to each other isn’t enough for contraction to occur.
Intransitive Want
There is an intransitive form of “want”, such as in the sentence
“I am never satisfied and always wanting”. This form also doesn’t
seem to contract.
- (In order) to be an effective over-consumer, you have to really want.
- You have to really want (in order) to be an effective over-consumer.
- You have to really WANNA be an effective over-consumer.
You could read (13) without the parentheses, and it would be ambiguous — (14), however, is suddenly unambiguous!
Parentheticals
There are also more “random” reasons why “want” and “to” might be
next to each other, such as in an interjection. In these cases,
contraction seems to be blocked as well.
- I want, to be precise, a yellow four-door De Ville convertible.
- I wanna, be precise, a yellow four-door De Ville convertible.
The old-school answer, that traces are the secret to wanna-contraction, just can’t be the whole story!
A Better Solution
Act 3 — An improved answer to the question “what are the rules for wanna?”
Those previous examples were noted by Pullum (1997), who developed a much more convincing solution to our conundrum: what if “wanna” is its own word?
It isn’t that extreme of a thought! It would certainly explain our data! “Wanna” would be its own verb (distinct from “want”), which would have its own (equally distinct) behaviors.5
And it’s not uncommon for phrases or contractions to become their own words (or lexicalize). Take “don’t” for example — native speakers would agree “do not” sounds way scarier than “don’t”. It doesn’t matter if we’re speaking slowly or quickly, they’re not exactly the same.
So, just as “do not” and “don’t” are subtly different, maybe “want to” and “wanna” are also different (though a bit more different).
The Verb “Wanna”
OK, let’s say “wanna” is its own verb… now we have a new problem! If “wanna” is a verb, shouldn’t it conjugate like one?
| Expanded | Contracted |
|---|---|
| I want to go! | I wanna go! |
| She wants to go! | She wannas go! |
| She wants to go! | She wansta go! |
Instead of “wanna” conjugating as if it were its own word, conjugation happens at the expanded level. Of course, “wanna” could just be irregular, but that seems a bit like cheating.
This is a problem for the theory that “wanna” is a new word. No fears, though — we can make a slight amendment!
The To-Postfix
Pullum makes a pretty clever jump: the secret is that “to” isn’t a word, but a morphological postfix! .
- Postfix
-
A postfix isn’t a word on its own, but something you may attach to
the end of a word to build a new word.
e.g. “Dog” + “-s” = “Dogs”
While “dogs” is clearly derived from “dog”, it has new behavior distinct from the singular “dog”.
The claim is that “want” may take the postfix “-to” and create a new word, with these altered grammatical properties! This allows for the conjugation pattern we saw before, while maintaining the theory that “wanna” is distinct from “want”.
We should expect the same thing to take place with other verbs as well — “to” should attach to some verbs, and modify their properties in a consistent way. And lo and behold, such a set of verbs exists: the therapy verbs!6
- want + to \(\Rightarrow\) wanna
- used + to \(\Rightarrow\) usta
- got + to \(\Rightarrow\) gotta
- going + to \(\Rightarrow\) gonna
- have + to \(\Rightarrow\) hafta
- supposed + to \(\Rightarrow\) supposta
And just as predicted, all of these words are imbued with new behaviors, different from that of the expanded forms. Let’s look at two examples.
“Going to” VS “Gonna”
- I’m not going to annoy her.
- I’m not gonna annoy her.
While (17) has two possible meanings, (18) only has one meaning: my full intention of annoying her. Only the “future” sense of the word may take the “-to” postfix (and not the locative).
“Have to” VS “Hafta”
Similarly, only one sense of “have” may take the “-to” postfix —
the necessity meaning.
- There are some puzzles I have to work on.
- There are some puzzles I hafta work on.
Again we see two possible readings for (19), but only one for (20). What’s more, “hafta” follows the same conjugation rules as “wanna”:
- There are some puzzles she haftas work on.
- There are some puzzles she hasta work on
Parting Notes
So is this the end-all-be-all answer to why “wanna” is such a strange beast? Of course not. I’ve completely skipped over the morphological rules that might allow “to” to behave this way, and they aren’t without controversy. But hopefully, if I’ve convinced you of anything, it’s that “wanna” (which we might have thought was just our tongues being lazy) is actually a beautiful and complex bit of language.7
If you’re looking for further reading, I highly recommend starting off with the Pullum paper -Pullum (1997). From there, you might be interested in some other work in the area.
- Boas looks at “wanna” from a Constructivist viewpoint, a totally different linguistic framework. It gives a good view of other ways to look at the problem (Boas 2004).
- There is research on how non-native speakers seem to learn these rules in ways different from native speakers (Kweon and Bley-Vroman 2011).
- Finally, an interesting paper by Zukowski and Larsen point out that children seem to use wanna in different ways than adults, and casts some doubt on modern thinking of “wanna” (Zukowski and Larsen 2011).
I’m curious what might have stood out to you? Are you a native speaker who didn’t agree with some of these judgments? Or maybe a non-native speaker who never noticed how complicated this contraction was? Maybe you’re aware of a similarly odd verb class in your own language! As always let me know your thoughts, and thanks for reading!
