He Wanna Be Adored

by Brian on January 30, 2004

This is a fairly rambling post on the syntax and semantics of ‘want to’ and ‘wanna’, so it’s almost all going below the fold. I would be interested to hear back if people agree or not with some of my judgments about the various cases.

Europa Malynicz sent me the following case, which it turns out we have different intuitions about.

Brian is running for election on the Monster Raving Loony Party ticket. He did not want to win when he entered, because he’s just smart enough to know that Monster Raving Loonies are bad for government. But, being a bit of a Monster Raving Loony himself, he’s now forgotten that he’s in the contest. While watching nightly news he sees himself dressed up as a Teletubby disguised as a teacup campaigning for votes. Not recognising who he is, but being very impressed with the hat he’s wearing, he forms the desire that that guy wins the election. In this context, which of (1) and (2) are true?

(1) Brian wants Brian to win the election.
(2) Brian wants to win the election.

I think that (1) is probably true, although it might be misleading. But I think (2) is false. (Europa thinks it is also true but misleading, which is where the difference arises.) So the first reason I’m posting this is to check what your intuitions are. Could (2) really be true in this context? Could (1)?

This is a cute case, and it suggests a couple of wild hypotheses.

De Se Hypothesis
(2) is only true if Brian has a de se desire, a desire that is essentially self-directed. It’s false in the case described because he has a de re desire that that guy wins, a desire that is directed at the guy on TV, which just happens to be him.

PRO Hypothesis
It’s really hard to explain the difference between (1) and (2), i.e. that one of them is true and the other false, if there is a hidden pronoun PRO between ‘wants’ and ‘to’ in (2). So here’s some evidence that there isn’t one. (I make no claims to originality here – debates about hidden pronouns are very well worked over so I suspect someone has made this point before.)

Thinking about PRO makes me think about wanna, but thinking about wanna just makes me confused. Here’s roughly why. How grammatical do each of the following seem to you?

(3) a. I want to be adored
      b. I wanna be adored
(4) a. You want to be adored
      b. You wanna be adored
(5) a. He wants to be adored
      b. He wanna be adored
(6) a. She wants to be adored
      b. She wanna be adored
(7) a. They want to be adored
      b. They wanna be adored

My intuitive judgments are that all the a sentences are OK, while 5b and 6b are very bad, and 7b is pretty bad. It turns out Google agrees that 5b and 6b are much worse than 3b and 4b, but it also thinks 7b is just as bad as 5b and 6b. Still, all the forms are used widely enough that they seem kinda acceptable at least to some. Here are the counts for searching for pronoun followed by “want to” or “wanna”.

want to

79,400 is many more hits than I expected for a phrasing I regard as marginal, an 28,300 between ‘he’ and ‘she’ is way more than I expected for a phrasing I regard as clearly bad. At least the ratios were way down than from ‘I’ and ‘You’.



Kieran Healy 01.30.04 at 6:52 am

Intuitions would be sharply divided were we to have “I wanna be sedated” instead of “I wanna be adored”. I predict that people who’ve recently turned 40 will favor the former regardless of case, and people who’ve recently turned 30 will favor the latter.


Brian Weatherson 01.30.04 at 7:03 am

Well, language changes all the time. The under 25 crowd probably all think “She wanna be admired” is well-formed English, or at least well-formed NewYorkEse.


--kip 01.30.04 at 7:16 am

Kids these days. Does no one say “He wantsa” and “she wantsa” anymore?

“He wanna” and “she wanna” just don’t register to my ear. “He don’t wanna” or “she don’t wanna,” sure, with the “don’t” taking the edge off the bad juxtaposition, but most aural elisions of third-person singular “wants to” honestly come out as “wantsa” to my ear. Of course, we’re talking about Google and typing, not ears and speech. —“They wanna,” though, seems no more or less unusual or malformed than “I wanna” or “you wanna.” For what it’s worth.


Peter Murphy 01.30.04 at 8:53 am

Afraid not, kip. “He wantsa” clocked a pitiful 311 hits on Google.

The odd thing about “wanna” is that it seems to work out better with nouns than pronouns. Compare “Polly wanna cracker” versus “He wanna cracker”. Of course, the first is a hard working clicle, but at least it works. The second doesn’t work at all.

Now I’ll have Stone Roses tunes going through my head for the end of the day. There are worse things.


raj 01.30.04 at 12:23 pm

It should be “wannabe,” not “‘wanna’ be” as a substitute for “want to be.” Not “wanna” be as a substitute for “want to” be.

BTW, if you believe that I’m kidding, I’m not.


reuben 01.30.04 at 1:12 pm

I wanna be sedated, but I want to be seduced.


wolfangel 01.30.04 at 1:57 pm

I wonder how many of the s/he wanna forms weren’t from s/he wants to but something like What does s/he want to do? Or anything where the form is just plain want.

7b sounds as good as 3b and 4b to me, but 5b and 6b sound awful.


John O'Neil 01.30.04 at 2:15 pm

Since I’m on the record on this issue, having written a PhD dissertation on the syntax of control clauses (“Means of Control”, Harvard U.), I think I probably should express an opinion on at least the semantics questions above. (As for the whole “wanna”-contraction debate, I’ve never been able to care enough about it to come up with a significantly better explanation than anyone else.)

Anyway, let’s assume that the difference between the control and non-control clauses is that, in the control clause “Brian wants to win the election”, the subject “Brian” is a direct argument of both the verb “wants” and the verb “win”; in other words, in some sense that’s dictated by one’s preferred syntactic theory, the argument positions are composed and then assigned to the same argument. However, in the non-control sentence “Brian wants Brian/himself to win the election”, we have the two argument positions being assigned to different arguments, admittedly with the same reference, at least to the first approximation.

However, if the embedded clause can refer to a different possible world (or a different discourse context, possibly), then it’s possible to imagine that you’re talking about different Brians in the non-control sentences — that is, the analogous Brians in different possible worlds. If this is an appropriate way to reflect the difference in knowledge between Brian and, well, Brian, then it also explains why the control clause doesn’t work in this case — because there aren’t two different “Brian” arguments to be in different possible worlds.

However, since I’m not a formal semanticist by any means, perhaps I’ll take this opportunity to stop digging the hole deeper.


Mario 01.30.04 at 2:19 pm

The odd thing about “wanna” is that it seems to work out better with nouns than pronouns. Compare “Polly wanna cracker” versus “He wanna cracker”. Of course, the first is a hard working clicle, but at least it works. The second doesn’t work at all.

I think the “Polly” example works not only because it is a cliche, but because it is the question, “Does Polly want a cracker?” not the statement “Polly wants a cracker.” As a statement, it has the same problem that the other third person statements have.


Sporty 01.30.04 at 2:20 pm

Raj is right…

“If you wannabe my lover?”

after all….


derPlau 01.30.04 at 2:40 pm

About half of the “he wanna”/’she wanna” searches on Google seem to be song lyrics — many of them repeats of the same song.

Doing a google search on “he wanna” but excluding pages with the words “lyric” or “lyrics” drops the result count from “about 12,000″ to “about 6,000″.

A glance at the first few pages of the full search shows that most of the song lyrics are repeats of songs by Michael Jackson (“Superfly Sister”), 2Pac (“Shorty Wanna Be a Thug”), Aaliyah (“Extra Smooth”), and Bukka White (“Sleepy Man Blues”). Excluding those songs (using the search “he wanna” -superfly -shorty -”sleepy man” -”extra smooth”) drops the total to “about 9,290″.

More or less the same thing happens with “she wanna”. This addresses my bizare obsessions far more than it does the subject of the original post, but it’s clear that the results from Google aren’t 12,000 independent uses of the phrase “he wanna”.


Scott Martens 01.30.04 at 3:08 pm

Brian, are you sure you’re not a closet lexicalist? (Or have you come out? I don’t know you that well.)

We’ll get rid of the reflexiveness in the question for a moment, to make things less ambiguous. Consider the argument structure in the following two sentences:

1 – John wants Peter to die.
2 – John wants to die.

Someone like Wierzbicka or Mel’cuk would define the verb in two different ways in these two cases:

1 – X wants Y to Z = X(cognitive agent) wants that subclause[Y(object) Z(action)]
2 – X wants to Z = X wants Y to Z = X(cognitive agent) wants that subclause[X Z(action)]

This confusion is therefore entirely lexical, and goes away if you consider words to only have meanings in conjunction with the structure of their arguments.

Our intuition that the problem is lexical is verified by translating these sentences into French:

3 – Jean veut que Pierre meure.
4 – Jean veut mourir.

Here, the subclausal nature of the underlying semantic construction in [1] is made into a surface syntactic feature, while in English it is absorbed into the structure “want + 3 arguments”.

So, what you are really seeing here is where two lexemes that are distinguished by their argument structure alone in Standard English are, in Vernacular English, increasingly distinguishable morphologically.

5 – John wants Peter to die.
6 – John wants to die.
7 – You want Peter to die.
8 – You wanna die.
*9- You wanna Peter to die.

This intuition can, once again be verified by French translation:

8 – You wanna die
10 – Tu veux mourir.

7 – You want Peter to die.
11 – Tu veux que Pierre meure.
*12- Tu veux Pierre mourir.

The surface word “wanna” is simply the first and second person singular and the universal plural form of “to want to” + 2 arguments. This is evidence that the two argument structures designate different verbs despite homophonous surface forms in Standard English.

Of course, this only makes sense in light of my judgement that your sentence 7b is fine, and otherwise I agree with your judgements.

Under this analysis “Brian wants to win” is inherently reflexive. It is by semantic necessity self-directed. I am inclined to prefer answer 1:

? Brian wants Brian to win the election.

I prefer this because the markedness of the sentence reflects the oddity of the situation. Brian does not want himself to win, he wants a person he doesn’t recognise, but who happens to be himself, to win. The semantic distance strongly implied by the “to want to + 3 args” lexeme fits this bill better, while the semantic identity required by the “to want to + 2 args” lexeme does not.


Scott Martens 01.30.04 at 3:18 pm

I should add that this can also account for the following sentences, but only if I push my case:

1 – He wants to be adored.
2 – *He wanna be adored.
3 – Does he wanna be adored?

The only way I can make sense of this is to assume that the infinitive form “to wanna” is also a part of the vernacular morphology of “to wanna + 2 args”. However, I must then have a special rule to account for this “to” that appears out of nowhere in sentence [1].

Since that would breach Occam’s Razor, I think the case I made above about the morphology of “to want to” is pure bollocks. It’s simpler to consider “wanna” a liason type phenomenon that occurs under any conditions where the argument structure places the surface forms “want” and “to” next to each other and leaves “to” a dependent surface form of “want.”

However, the notion that there are two lexemes here, that are distinguished by their argument structures, remains sound. As, I believe, does my answer to the original problem.


Scott Martens 01.30.04 at 3:19 pm

…leaves “to” a dependent surface form of “want.”

should read:

…leaves “to” a surface form dependent on “want”.

I think I had better get back to work now.


Matt Weiner 01.30.04 at 4:19 pm

Kieran is wrong–I’m 32 and I’ll take “I wanna be sedated” any day. (Er, unless he doesn’t count 2001 as recent…)

I think “wannabe” is a noun or adjective, while “wanna be” is a verb. (If you think I’m kidding, you’re mostly right.)


JRoth 01.30.04 at 4:52 pm

I can’t attach it directly to anything anyone else has said, but surely it’s significant that, in January of 1964, John Lennon and Paul McCartney wanted to hold the hand of English girls, but wanna’d hold the hand of American girls.

If that was too awkward, I’m simply saying that the English release of that song was titled “I want to hold…” and the American release was “I wanna hold…”


msg 01.30.04 at 7:38 pm

Kip gits it.
Only “wantsta”. People have not been given permission to write “wantsta”. “wanna” they have.
It’s a narcissism thing with “wanna”. The infantilization of the polis.
“Wantsta” is third-person, which means a recognition of the other.
“Wanna” is threshold speech. Comes right after “yes” and “no” in the language armamentarium.


triticale 01.31.04 at 2:21 am

Does Mary wanna smoke some marywanna?


seth edenbaum 01.31.04 at 5:58 am

Wanna = want
But there’s no law, and no accounting for taste. In any event it’s not a question of right or wrong but of unnecessary ‘uneconomical’ complexity. But poetry or lyrics are not meant to be economical. So why not add the extra baggage? It adds a layer of ironic distance.
T.S. Eliot would approve.


seth edenbaum 01.31.04 at 5:59 am

Generally speaking
Wanna = want
But there’s no law, and no accounting for taste. In any event it’s not a question of right or wrong but of unnecessary ‘uneconomical’ complexity. But poetry or lyrics are not meant to be economical. Why not add the extra baggage? It adds a layer of ironic distance.
T.S. Eliot would approve.


s. e. 01.31.04 at 4:02 pm

One more comment. Wanna is a contraction of ‘want to’, and replaces it in common usage without changing the meaning. It performs an identical function. It’s simply easier to speak the transition from vowel sound to vowel sound ‘a’ to ‘o’ without the ‘t.’ From there the transformation of the second vowel ‘o’ to the a repeat of the first producing a-n-a is perfectly logical. Less energy expended, same result.

The use of ‘wanna’ for ‘wants to’ not only simplifies the process of speaking, but simplifies the language itself. The word now performs two functions, as ‘want to’ OR ‘wants to,’ and the result is a ‘contraction’ not just of sound but of meaning, and that denotes a vulgarization.

Comments on this entry are closed.