What the F Page 12
So is it possible the pope’s error was due to a mere planning inefficiency? Any single instance of a speech error could have many causes, and the only way we can tell that planning plays a role is in aggregate across many errors where the patterns reveal themselves statistically. But if we closely examine exactly what the pope said and what he was planning to say later, we can determine at the very least whether planning ahead is a plausible explanation for his caso to cazzo slip. So here’s the full text of the sentence containing the error, as released by the Vatican in the official transcript of his prepared remarks. WARNING: The following may contain Italian.
Se ognuno di noi non accumula ricchezze soltanto per sé ma le mette al servizio degli altri, in questo caso la Provvidenza di Dio si rende visibile in questo gesto di solidarietà.8
In this critical sentence, the pope is making an appeal for charity, as you can tell from the English translation: “If each one of us does not amass riches only for oneself, but half for the service of others, in this case the providence of God will become visible through this gesture of solidarity.”
Let’s look at the words that followed caso for possible anticipation candidates. There’s la, the feminine version of the definite article “the,” and then Provvidenza (“providence”). Notice that Provvidenza has a z right before the last vowel, just as caso has an s right before its last vowel. Could Francis have been planning the long, complicated word Provvidenza, while still articulating caso, such that he replaced the s of caso with an anticipated z? It’s possible. The two sounds are quite similar,b and they’re in the same place only two words apart, in words that are also of the same part of speech (which can increase the likelihood of errors).9 But this close reading of the text doesn’t lead to anything conclusive. At best, it tells us we should hold on to preplanning as a reasonable suspect.
But they’re probably not the sounds you’re thinking of. In Italian, the letter s in caso is pronounced how an English speaker would pronounce z, and the Italian z in Provvidenza is pronounced like an English ts. The double zz of cazzo is pronounced as an elongated ts. But this doesn’t substantively change the argument I’m making; it just makes everything more complicated if you don’t speak Italian, which makes it seem like the type of thing a considerate author would quarantine in a footnote.
There are, of course, other potential causes of the pope’s error. The popular conception of speech errors holds that they’re due more to meaning than mechanics. Sigmund Freud famously argued that when you make a speech error, it can reveal the inner workings of your unconscious mind.10 He wrote, “Almost invariably I discover a disturbing influence from something outside of the intended speech . . . a single unconscious thought, which comes to light through the special blunder.”
Certainly Freudian slips abound. When Tiger Woods had to withdraw from a tournament with a sore neck in 2010, only a few months after a dozen women had accused him of infidelity, a reporter on the Golf Channel announced that Woods was suffering from a bulging dick.11 Freud might interpret this error as deriving from something the speaker was actually thinking about, which managed to sneak its way into her words. Another well-known slip is credited to Condoleezza Rice, then national security advisor under President George W. Bush. Rice was famously committed to her job (as of the time of this writing, she has never married or had children), so it caught the White House press corps off guard when, at a Washington, DC, dinner party, she was reportedly heard to say, “As I was telling my husb- . . . As I was telling President Bush.”12 The word husband doesn’t sound much at all like President Bush, so the best explanation is that rather than sound similarity between the words, some aspect of meaning—what she intended to say or what she was thinking about—drove this error. But we will have to leave determining the precise semantic motivation in this case to those more familiar with the principles of psychoanalysis.c
This reminds me of a joke about Freudian slips: A patient tells his doctor, “Doctor, last night I made a Freudian slip. I was having dinner with my mother, and I wanted to say, ‘Could you please pass the butter.’ But instead I said, ‘You manipulative bitch, you completely ruined my life.’”
Could there have been similar Freudian motivation for the pope’s articulatory malfunction? Maybe. Seventy-five years after Freud’s death, the principles of Freudian psychoanalysis have largely fallen out of favor, at least among most researchers interested in cognition and behavior. But remnants of his theories persist, particularly when they aid in predicting behavior. And taboo language shines in speech error experiments based on Freudian premises.
Suppose, as Freud would have it, that whether you’re the head of the National Security Agency, the pope, or just you, things you’re thinking about but not intending to say out loud influence the speech errors you make. How could you tell? You might come up with an experimental paradigm like the one used in a set of studies conducted by psychologist Michael Motley and his colleagues in the late 1970s and early 1980s. They had people come into their lab and read pairs of words out loud, pairs like back mud, bad mouth, and so on. In using this paradigm, they found that people make errors at a certain base rate. And they wanted to know whether that rate would go up when people were thinking about something that they were actively trying not to talk about. So here’s the methodology they devised.13 It’s pretty clever.
Participants were all young, self-identified heterosexual men. Half were greeted by Motley himself, dressed professionally, and the other half were met by a provocatively dressed, young, female research assistant. As Motley describes it, “She was wearing a translucent, nearly transparent, off-the-shoulders top with a super-short yellow skirt.” Motley continues, “And we had her sitting on a stool where her knees were at eye-level with the guys.”14 It goes without saying what the young, heterosexual, male participants were expected to be thinking about.
In the experiment itself, participants had to read word pairs at a rate of one pair per second—a pace similar to normal speech. And as the critical component of the experiment, there were two types of word pairs. The first included pairs like mad bug. As you can see, making an anticipatory or exchange error in pronouncing this pair would produce something totally innocuous, like bad mug. But the second type would produce sex-related errors—let gaid makes get laid and share boulders gives you bare shoulders.
When they counted the number of exchange errors the participants made, Motley and colleagues found that the half of participants who had been sitting with the scantily dressed research assistant did indeed make more errors overall. But the increase was driven entirely by the sex-related word pairs: participants made more errors on let gaid but not on mad bug. They made the same number of innocuous errors regardless of who was sitting next to them.
We might never resolve whether this means, as Freud would explain it, that our cognitive unconscious is straining to have its say through daily speech errors or simply that when you’re thinking about sex, you’re more likely to say sex-related words. But it is clear that things you’re thinking can make their way to the surface, sometimes overcoming your will to suppress them. This tells us that selecting words to articulate is more complicated than merely picking the right words for the meanings you want to express. It involves selecting and suppressing thoughts as well—because under certain conditions, those thoughts bubble up in the form of speech errors.
As a mere fallible human—and possibly a heterosexual male one to boot—even the pope is not immune to this sort of Freudian influence. In inadvertently uttering cazzo (“dick”), could the pope have been talking about the virtues of charity but thinking about his vow of chastity? It’s possible.
We’ve established that the pope might have erred due to the sequence of sounds he was planning to utter, and we’ve entertained the Freudian possibility that things underneath the vestments were on his mind. But he was also speaking under challenging conditions. This wasn’t an oration he delivered alone in the shower. Public speaking can be distracting—
feedback from the amplification system and things happening in the crowd draw your attention from the task at hand. And it can be stressful too. Both of these factors increase the rate of speech errors.15 I have occasion to observe this frequently in daily life. I live in San Diego, and despite the reputation America’s Finest City enjoys for outstanding news broadcasting,d the local announcers are a bit uneven. One particular afternoon host on the public radio station has a habit of stumbling over words, whether describing the upcoming segment on All Things Considered or announcing the names of companies providing local underwriting. And there have been some doozies, as I’m sure you can imagine, when the local businesses in question have names like Chism Brothers Painting and Bastyr University.e The pressures of public address must surely be as challenging for popes as for anyone.
For a historical perspective, see the following documentary: Apatow, J., and McKay, A. (2004).
Paradoxically, we hold people to a higher standard in exactly those conditions that are most likely to induce errors. Even knowing as I do that speech errors are an inevitable part of speech production, especially when a person is experiencing stress (as radio broadcasters probably do), I often find myself yelling at the radio, “Come on, step up your game! This is public radio! You don’t think I donated just for the tote bag, do you?” And then I remember that I didn’t donate this year, and I feel remorseful, but then I rationalize not donating with the thought that if the broadcasters didn’t bungle their delivery all the time, maybe they’d deserve my money. The human mind is a silly place.
What’s more, the pope was speaking in a foreign language, which makes fluent speech harder. You make errors in a foreign language simply because you don’t know the gender of a word or because you have a tenuous grasp of some detail of the language’s grammar. (Do indirect object pronouns come before or after the verb? And for that matter, what’s an indirect object pronoun again?) These lack-of-knowledge-based errors aside, your base rate of normal slips of the tongue also goes up—by 1400 percent.16 So perhaps we shouldn’t be surprised that delivering an address in Italian, instead of his native Argentinian Spanish, might ratchet up the frequency of a pope’s errors.
It seems reasonable that a fallible pope ought to be subject to the same pressures and linguistic traps as anyone else, and in this particular case they may have conspired to generate the linguistic C-bomb we witnessed. Indeed, with all these pressures at play, it’s surprising that what comes out of your mouth—or the pope’s—isn’t just a stream of mistakes. Strangely, for the most part, it isn’t. While we all make speech errors, the majority of words we produce really are precisely what we intend. And as we’ll see, profanity provides the most revealing clues to how we accomplish this.
# $ % !
Some psycholinguists have hypothesized that the only way you could possibly be as good at speaking as you are is by somehow monitoring your planned speech. You might, in this view, have an internal editor in your head that pays attention to the words you’re planning to say, the order you’re planning to say them in, and exactly how you’re planning to pronounce them. It’s like quality control at the end of the assembly line, right before the words get packed up and leave the factory. When your internal editor notices something about to go awry, it stops the conveyer belt and sends the offending word back for repair. Of course, some errors get through, so we know the editor can’t be perfect, but the idea is that perhaps internal self-correction keeps your errors down to the acceptable level they’re at.
How would we go about detecting such an internal editor? It’s tricky because an editor would leave little trace. If there is indeed an editor, and if it’s mostly successful, then there will be a few but not many errors to observe. Likewise, if there isn’t an editor, there will be a few errors. The problem is figuring out how many errors a person would have made without an editor if we don’t know whether there’s an editor involved in the first place.
Let me flesh this out with our factory analogy. Suppose you want to know if a factory that cans diet soda has a quality control department. You can start by observing errors—every once in a while, someone finds a cockroach in a sealed can of diet soda. Let’s say it’s once out of every hundred million cans opened. Now, were thousands of other roaches in cans of soda caught by a vigilant quality control department? Or is the fabrication process itself so hygienic that an interloping roach finds itself trapped in a saccharine sarcophagus only one time in one hundred million? How would you tease apart these alternatives? Seems like a dead end (and not just for the cockroach).
But here’s a possible way forward. To continue with the analogy, suppose you know that a cockroach finding its way into certain types of soda would be particularly devastating for the factory’s reputation. For instance, suppose that the same factory packages the very same liquid not just as a brand-name soda but also, labeled differently, as an in-house, generic supermarket label. The only difference is the cans. And let’s presume that the company has a greater incentive to ensure that the brand-name version is roach-free because it’s a bigger source of revenue, and a single photo of just one roach in the brand-name soda will break the Internet and gut the company’s bottom line. By contrast, the company might reason that people sort of half expect to find insects in in-house supermarket-label soda. Maybe that’s even why they buy it. Here’s the point. If there’s no quality control department, then you’d expect to find cockroaches in the brand-name and generic sodas with about the same frequency; they’re produced and canned in the same factory using the same process. But if you found far fewer canned roaches in the brand-name versus the generic soda, that would tell you that someone’s making sure that when it matters more, mistakes don’t make it out the door and onto the truck.
Several psycholinguists have used exactly the same logic with speech errors. To do this, you need to find certain speech errors that would have graver consequences than others. What’s the linguistic equivalent of a cockroach in a brand-name can of soda? Well, the speech errors with the direst results are probably those that generate profanity. So the question becomes, when you put people in a position to say the wrong words, do they make the same number of errors, regardless of whether the error would produce profanity? If so, then there’s no evidence of an internal editor. But if people make fewer profane mistakes than nonprofane mistakes, then that implies that people are internally suppressing the errors before they hit the tongue. They’re self-monitoring language.
If you want to use this logic, you have to devise a way to induce speech errors in the lab. The first group to do this came up with a clever design.17 You’ll recall Michael Motley, the researcher with the provocatively clad research assistant in the Freudian slip study. He and his colleagues had people read carefully designed word pairs one at a time. Some of the pairs had the potential to become obscene if the participant made an exchange error. For example, tool kits seems totally innocuous, until you recognize how it would sound as a spoonerism. You can construct lots of potentially obscene lures like this: bunt call, hit shed, duck fate, heap chore, fast luck, and so on. The key question in Motley’s studies wasn’t whether you can get people to mistakenly mispronounce these. You can. Following the cockroach logic, the question was whether people would make fewer errors on profane-potential pairs like these than on pairs that do not threaten obscenity. So there was a second type of pair, like tool kicks. These words are identical in most every way to tool kits—they start with the same sounds, they’re the same length when pronounced, they’re the same parts of speech, and so on. The difference is that the errors you might make, producing things like cool ticks, aren’t in the least taboo. Would there be more errors on these inoffensive pairs than on the offensive ones?
One methodological note, in case you’re interested in trying this at home: Just reading pairs of words out loud doesn’t yield many errors. So you need to boost the error signal. One thing you can do (and the researchers did) is stack word pairs one after another leading up to th
e critical one in order to set people up for failure. For instance, if you want people to make an exchange error on tool kits—swapping the initial t and k—you’d stack pairs in front of it as below. Try reading these aloud:
kind tiger
calm time
cold tea
tool kits
Setting people up with the swapped consonants before critical pairs like tool kits or tin cable makes participants much more likely to produce errors. This provides more opportunities for mistakes, which makes potential differences in the frequency of taboo and nontaboo errors easier to measure.
So, all other things being equal, do people make fewer errors on pairs like tool kits, where the result would be offensive, than on ones like tin cable, where it wouldn’t be?
Two studies did this originally, in 1981 and 1982. The chart you see on the next page shows the number of errors people produced on average. There were many more neutral errors than taboo ones in both studies (though the difference was bigger in the second study). It follows that people in these studies were successfully avoiding errors specifically when the results would be obscene. Fewer roaches in the brand-name soda. People were self-monitoring.
But this is only the beginning of the evidence. If you’re clever enough, as the researchers working on this are, you can come up with some other things you’d expect to see if people were doing internal quality control. Here’s one. Suppose you’re on quality control at the factory and you find a can with an unwelcome passenger inside. Sending it back to be fixed or replaced should add time to the process. So even when the ultimate product shows no sign of error, the time it takes to produce it could be a hallmark of monitoring. Mapping this over to speech, if the lower error rates with taboo word pairs are due to editing, then that should show up on how long it takes people to produce the pairs correctly—it should take people longer to say pairs correctly when they’ve planned an error but subsequently taken the time to catch and correct it.