What the F
Page 22
According to this Scandinavian survey, profanity is not demonstrably one of those types of abuse. Studies like this one don’t tell us anything about profanity per se because all such studies slip profanity in with things that are clearly verbal abuse and don’t allow us to tease them apart. In this case, profanity was lumped together with things like threatening with violence, throwing objects, and insulting the child, for instance. And any participant who said yes to any of the verbal abuse questions was categorized as having been subject to verbal abuse. So there’s no way to tell whether these other activities—which clearly constitute abuse—were responsible for the entirety of the effects observed or whether profanity also played a role.
And there’s good reason to suspect that if you were to split out exposure to profanity, you wouldn’t see any effect on child well-being at all. Profanity can, of course, be used for many different purposes. It’s absolutely true that some of those are abusive. But it stands to reason that calling a child worthless or threatening physical harm is likely to increase the child’s depression and hostility, whether profanity makes an appearance or not. Conversely, it’s hard to imagine what harm is likely to come from telling an adolescent his report card is fucking fantastic. Compare that with telling him You’re stupid, and I hate you. You can use profanity positively, and you can use “clean” language abusively. Profanity is orthogonal to abuse.
Verbal abuse of children is, of course, an issue we should be studying more and raising vigilance about. But there’s no evidence either that profanity itself does direct harm to children or that abusive language is more harmful when it includes profanities. Again, I have to raise the specter of the absence of evidence. We know only that we don’t have evidence that profanity hurts children directly. But in the absence of that evidence, one would also expect an absence of the pretense that evidence exists.
# $ % !
So we’ve debunked the Pediatrics article’s first claim: that exposure to profanity harms children directly. Let’s move on to the second claim: that the more children are exposed to profanity, the more they will come to use it themselves.
In the limit, a categorical version of this claim must be true. Children learn language based on what they hear around them. That usually includes profanity. So if they don’t ever hear profanity, they won’t learn it. And although many parents try to shield their children from profanity, this is largely impossible. Parents often try to self-censor when a child arrives, but this can be challenging, as anyone who has stepped on a Lego block in bare feet or been micturated upon at the changing table can attest. Not every swear gets stifled.
And, of course, children learn language not just from their primary caregivers but also from other speakers of the language around them. And these other people often don’t buy into the idea—an aunt, grandfather, older sibling, or friend might not see any problem with an occasional shit here or bastard there. Maybe one of those individuals even secretly—or not-so-secretly—wants to sabotage the whole “linguistic whitewashing for the children’s sake” enterprise. So children are going to be exposed to profanity by people around them.
And also: the Internet.
But the article proposes a more nuanced, quantitative relationship. Does more frequent use of profane words around a child lead that child to use them more? Maybe.
There’s a long history of studying how a word’s frequency in a child’s environment affects acquisition of that word. Children typically begin to understand a few words in their first year of life and utter their first words around their first birthday, as we saw in the last chapter. From there, they’ll typically acquire a few words per month until sometime around eighteen months (give or take six months), when their vocabularies begin to explode in a “word spurt.” By the time they start school, their vocabulary will have ballooned to something like 10,000 words, and by college, it will be five times larger.
And it’s reasonable to think that the words children hear most frequently are also those that they learn earliest and use most. But it’s complicated. Children’s early words tend to name people, animals, things, parts of things (especially body parts), and actions, or to mark social interaction routines (like bye-bye, up, or all gone). The distribution of these sets of words in a child’s early lexicon varies between children and across languages and cultures—English-speaking kids learn more early nouns than their Korean-speaking counterparts, for instance.18 And while many of a child’s first words are frequently occurring ones (the name of a family pet or a word for a commonly experienced object, like a doll, ball, or bottle), a child’s earliest words often do not include the most frequent words in the language. For example, the ten most frequent English words (as measured by the spoken component of the American National Corpus)19 are, starting with the most frequent, the, of, and, to, a, in, it, is, for, and I. Yet none of these are commonly found in a child’s earliest words. In fact, if you look at caregiver speech directed at children (“child-directed speech”), you find that more frequent words are learned slightly later than less frequent ones, in large part because more frequent types of words (like articles, prepositions, and so on) are learned later than less frequent words like nouns.20
So at the very least, we can say that the story is more complicated than merely frequency influencing word learning. In fact, when you split words into different groups—common nouns, words describing people, action words, and so on—you do indeed find that within each category the more frequent words tend to be learned earlier. For example, here’s a graph of the age at which one particular child first used each of his nouns (his age is on the x-axis) plotted against how frequent that word was in the child-directed speech he heard (it’s actually the log of word frequency because frequency effects in language have logarithmic effects).21 You can see that within nouns, the child learns more frequent ones earlier, on average, and then moves on to learn less frequent ones as well.
Each dot represents the first time the child produced a particular noun; more frequent nouns tended to be learned earlier than less frequent ones. Image reproduced from B. C. Roy et al. (2009), used with permission.
Of course, a reasonable person could object to studies like this one. Correlation does not imply causation. So the fact that children tend to learn more frequent words earlier doesn’t entail that frequency is the reason for earlier word learning. Other factors might be in play. For instance, more frequent words are shorter, all things being equal. And children learn shorter words earlier. Maybe frequency plays no causal role.
To know for sure, you’d need to run an experiment: you’d have to manipulate how often children heard particular words and see whether this factor alone, holding all other possible causes constant, affected children’s learning of the words. There have been several such studies. Perhaps the best known took place in the early 1980s.22 Researchers interacted with one- to two-year-old children over the course of four months and, among other things, introduced them to some made-up words, like tete or fus. The words had to be made-up because otherwise the children might encounter them by chance in their normal lives as well. And the researchers carefully manipulated whether, over those four months, the children heard each word ten times or twenty times. And they measured how likely the children were to spontaneously use those words. The result was clear: they learned the frequent words relatively often (on average, kids learned 44 percent of them) and far better than the infrequent ones (29 percent). This effect was observed for both nouns and verbs.
So it seems that a causal relationship exists between how frequent words are in a child’s environment and when he or she will start using them. But how well does this scale? Once an older child knows a larger, more adultlike number of words, say on the scale of tens of thousands, does the frequency with which those words entered the child’s ears affect the frequency with which they come out of his or her mouth?
We don’t have a conclusive answer to this question. We don’t even know exactly what proportion of children
know what profane words at what age, because that research hasn’t been done. Typical surveys asking parents about their children’s vocabulary tend to leave out profanity, and parents usually aren’t particularly forthcoming in offering swearwords that their children know. The one source we have is an observational study, in which a set of researchers spent a year recording all taboo words uttered by children around them—predominantly middle-class kids from the northeastern United States.23 The results showed that the swearwords children use change with age. The youngest English-speaking children are more likely to use words like poopy or stupid with the force of profanity, whereas by adolescence they’re using more adultlike profanity: fuck, shit, and so on. We also know that boys and girls swear differently. Oh my God ranks as the most frequent profane expression among girls aged one through twelve, whereas boys in the same range are most likely to use shit.
So we have some hint as to when children use which profane words, but we certainly have no evidence of how frequently they use them. The only exception is the Pediatrics study. And that evidence is rather weak, as we’ll see.
# $ % !
Let’s look at exactly what the Pediatrics study did. The researchers asked students from a large midwestern middle school a series of questions. To get a measure of how much the children were exposed to profanity, they asked the kids to list their three favorite television programs and video games, then rated these for how much profanity they contained. To determine what the children thought about profanity, the researchers asked them to give their opinion about statements like “I think it is okay for me to use profanity in my conversations” on a scale from 1 (never true) to 5 (almost always true). They also wanted to know how often the children actually used profanity, and to measure this they asked them to rate a series of statements like “I use profanity in my conversations with my friends” from 1 (never true) to 5 (almost always true). And finally they wanted to know how aggressive the children were. So they asked them to rate a number of statements describing physical aggression (“I hit, kick, or punch others”) or relational aggression (“I have tried to damage [a] person’s reputation by gossiping about that person”), again on a scale of 1 to 5.
With these data secured, the researchers dug into them to discover any significant correlations among the various measures. And some of them did correlate. In particular, children who reported being exposed to media with more profanity were also more likely to report that they found profanity acceptable. In turn, those who found profanity more acceptable were also more likely to report using profanity. And those who reported using more profanity were also more likely to report acting aggressively. The article interprets this chain of correlations as implicating profanity in aggression: observing profanity affects attitudes about words, attitudes affect the use of those words, and word use affects aggression.
But not so fast. Can a study like this actually tell you whether profanity causes this chain of effects?
Let’s tease them apart, starting with the first couple of steps. Right now we’re focused on the question of whether exposure to profanity increases its use. The study states that adolescents who reported watching shows and playing games with more profanity in them also reported finding profanity more acceptable and using more profanity themselves. Does this answer the question about frequency? Does this mean that exposure to more profanity leads to more use of profanity?
We don’t know, because the study was correlational. It’s not always obvious why correlation doesn’t imply causation, so let me just remind you here. (If this is old hat to you, by all means, skip to the next paragraph.) Here’s a nice example of why you can’t infer causation from correlation.24 Suppose you want to know whether religious faith causes an increase in alcohol consumption. You might try to find an answer by counting the number of bars and the number of churches in each of a large number of US cities. You’d surely find a significant correlation; overall, cities with more bars have more churches. But what’s the right causal interpretation of these data? That religious practice causes drinking? Maybe. I certainly know people who claim to need a drink after church. But the reverse causality is equally possible: maybe increased drinking causes more religious practice. You can come up with a story for this too. Maybe when people drink more, they do more things that they feel they need to atone for. And even worse, it’s equally possible that something totally unrelated to churches or bars—some hidden variable that you didn’t even consider—causes concurrent increases in both. For example, maybe cities with more churches also have more bars because those cities have larger populations in general. In short, a correlation between two things cannot tell us whether they’re causally related or, if they are, in which direction. That’s true whether we’re interested in the number of churches or exposure to profanity.
So let’s apply this reasoning to the Pediatrics study. The data about exposure to profanity, attitudes toward profanity, and use of profanity were all self-reported. So we know that children who reported exposure to media with more profanity also reported more positive attitudes toward using profanity, and in turn those who reported more positive attitudes toward profanity also reported using more profanity. You might be able to spot the hidden variables here. The problem is that all the data rests on the children’s reporting. We don’t have any idea of their actual behaviors, beliefs, or viewing patterns. And these children’s reports could differ from reality for many reasons. Perhaps some of the children intentionally misled the researchers. For example, some children could have been afraid to confess that they were exposed to, had positive attitudes toward, and tended to use profanity. Imagine yourself as a middle school student. A team of researchers from prestigious Brigham Young University shows up asking about profanity and aggression. You might think twice before giving truthful answers. Some children might be more honest than others. And if the less honest children artificially deflated their scores on each of these questions, this would produce the precise correlation that the study found. It would create the false impression of correlations in experience and behavior where the real effect is just a tendency for some children to respond less truthfully across the board. Note that it could also work in the other direction: some children could have been boasting about both the shows they liked to watch and the amount they swore. This too would create a spurious correlation in exactly the same direction: children who reported watching more profane media would boast of more positive attitudes toward profanity and more profanity use.
This is just one of many possible hidden variables that might describe the actual causes underlying the correlations that the study observed. For instance, maybe children with caregivers who don’t believe there’s anything wrong with swearing both allow their children to watch media with profanity and are more likely to encourage or permit its use.
And there’s also the problem of reverse causation to deal with. It could very well be, as the article suggests, that exposure to profanity causes increased use of profanity. But what about the opposite direction? Maybe children who swear are more strongly attracted to profane television shows because such programming sounds more like what they sound like or what they want to sound like.
So does increased exposure to profanity lead to greater use of profanity? The noun and verb research discussed earlier suggests that it’s certainly possible that children exposed to more profanity will learn it earlier. But that doesn’t imply that they’ll use it more, and the correlational study we’ve just been discussing is causal Swiss cheese. However, just to see where the argument goes, let’s suspend disbelief and stipulate, for the moment, the possibility that children will use profanity more the more often they’re exposed to it. The really important part of the story is whether using more profanity causes harm to children. That’s next.
# $ % !
The Pediatrics study discovered a correlation not just between profanity exposure and use but also between how much children reported using profanity and how much they reported engagi
ng in both physical and relational aggression. This part of the study suffers from the very same problems we just identified. For one thing, there’s still the hidden variable caused by self-reporting. As we saw before, correlations between self-reported variables might tell you more about how the respondents deal with the survey than about the behavior it asks them about. People in a study like this might not tell the truth for a variety of reasons. Adolescents who were more candid about aggression might also have been more candid about the media they were exposed to.
Many other possible hidden variables could be behind the correlations between profanity use and aggression. Which of these seem plausible to you? Maybe children exposed to more profanity also happen to be exposed to more violence in media, which causes them to be more aggressive. Or maybe children exposed to more violence in real life both seek out media with more profanity and also act out more aggressively. Or maybe children who come from families that are more predisposed to aggressive behavior for any of a number of reasons (medical, cultural, socioeconomic, genetic, and so on) both seek out more profane media and also exhibit more aggression. I’m not committing to any of these as the single, real cause of aggressiveness. But if you think any of these are plausible, and actually even if you don’t, they reveal the hidden variable problem inherent in correlational studies. People come with baggage. And some of that baggage might be causally related in a systematic way to the things you’re interested in.