Picture a Thursday dinner, or one close to it.
It is 6:40. The pasta is a little overdone, and nobody minds. Your daughter is twelve. She has just asked, for the third time this week, why her phone has to sleep in the kitchen.
“Everyone else keeps theirs.”
You say the rule is the rule.
“That’s not a reason. You don’t even have a reason.”
You do have reasons. You have several. But it is 6:40 on a Thursday, and you can hear your voice getting ready to say something your own mother used to say. So you say it. Because I said so. Her chair scrapes. Upstairs, her door closes, not quite a slam.
Later, wiping the counter, you find her English folder under the mail. Inside is an essay, a page and a half, on whether zoos should exist. She is against them. She says so in the first paragraph, again in the second, and a third time at the end. In the margin, in green ink, her teacher has written one line:
What would someone who disagrees say?
You stand there with the sponge in your hand.
Forty minutes ago, she argued with you without stopping for breath. On paper, she cannot picture anyone disagreeing with her at all.
To be fair to you both, at dinner neither of you was picturing the other side either. It had been a long day. Everyone was tired.
It looks like two problems: a child who talks back, and a child whose essays don’t. It may be one.
The skill nobody has a quiet hour for
At that table you were doing two jobs at once.
One is keeping the peace in a house with a twelve-year-old in it. The other is raising a person who will one day need to disagree with a boss, a doctor or a friend, and do it well: give her reasons, hear theirs, and change her mind if theirs are better. On a weeknight, those two jobs pull in opposite directions.
Many of us were never shown how to do the second job with a child. Many of us learned to argue the way she was arguing tonight: about the rules, with the person who makes the rules, at the end of a long day. That may be about the hardest place there is to practice anything. The stakes feel high, everyone is tired, and the same side usually wins.
You might hope school is where the other kind of argument gets practiced. Researchers have gone looking for it.
In the late 1980s, a research team sat in on more than 200 American eighth- and ninth-grade English and social studies classes, 872 visits in all, and timed how much of each class was open discussion. On average, it came to less than a minute per class period. In eighth-grade English it was about 50 seconds. In ninth-grade English it was under 15 seconds. In the lowest-track classes, it almost never happened at all. Of the things the team measured, the thing most closely tied to a discussion starting was a student’s own question.
Why would discussion like that be rare? Here is our reading; the study didn’t test it, and its timings are decades old. Picture one adult, a roomful of children, a class period, and a list of things to cover before the bell. There, the question with a known answer is the tool that works. It checks who is following. It keeps a whole room moving together. It treats every child the same. A real disagreement, the kind where one child’s reasons get answered by another child’s, takes time a schedule like that rarely leaves. So a student could go through a school year answering questions and spend very little time disagreeing with reasons, out loud, with someone who answers back. Then she is handed an essay prompt and asked to do on paper what she has had little chance to practice in the air.
That is not a failing of the teachers in those rooms. A room like that is built to cover the list.
And on Thursday, the one argument she had that night was about the rules, with the tired adult who made them.
The belief at the table
Before going further, the belief that ended the conversation deserves its full say.
Arguing is a bad habit. Good kids don’t talk back.
Many of us grew up with some version of that sentence, and it protects things worth keeping. A child who respects the adults in her house. A home where the loudest voice doesn’t win. A parent who doesn’t have to put every rule on trial before bed. Sass is not reasoning. Neither is showing off, or wearing someone down until they give in. If that is what “arguing” means in your house, the belief is right to stand in its way.
Listen to what the sentence means when a tired parent says it. Don’t talk back often means I can’t be on trial tonight. Sometimes it means I need you to trust that I have a reason, even when I’m too worn out to give it. And sometimes it means if I let her argue, she’ll just dig in deeper and get better at getting her own way. You have seen the digging in already, on a page and a half about zoos that says the same thing three times. Those are fair things to need, and a fair thing to fear.
Part of that last worry, the one-sidedness, has research behind it.
One review of developmental research argues that even young children can give reasons and weigh them, and can gain from reasoning with others. But it adds a catch. Like adults, children tend at first to argue only for their own side. On where she starts, the belief is right.
Researchers who study classroom group talk add a second danger: someone has to deal with status, so that the most confident voice (at a family table, we’d guess, the oldest child or the parent) doesn’t become the only one.
So nothing here asks you to reopen the phone rule every night. Some rules are not up for debate, and a child is better off knowing which ones they are.
What the belief can’t tell you is where the other kind of arguing gets learned: the kind with reasons in it, and room for the other side. If it isn’t learned in a fight about the rules, and only rarely in the class periods those researchers timed, then where?
Someone waiting for the reply
You have probably watched the other kind of argument between siblings: who had the controller first, with both children digging in harder every round. Nobody in that fight is looking for the other side’s reasons.
What an argument is for seems to matter. In one study, people argued with a partner, either to persuade or to reach agreement. Later, they wrote essays. The ones who had argued to reach agreement more often took the other side seriously on paper. They brought up claims that challenged their own position, worked those claims into it, and used points their partner had raised first. The researchers cautioned against teaching argument mainly as persuasion. Two children fighting over a controller are practicing something, but it is not this.
Something may also go inside. In a small British study, 8- and 9-year-olds were taught to reason out loud together: give your reason, and ask the others for theirs. The children got better at reasoning as a group, which you might expect. They also improved on a nonverbal reasoning test they took alone. Writing about this line of work later, researchers from the same program offered a possible explanation: the children may have learned to carry on a silent, reasoned conversation with themselves. They are careful to add that it is not clear exactly what the children learned.
Think of the green ink on the zoo essay. What would someone who disagrees say? is one thing written in a margin. It is another asked across a table, by someone who is waiting for the reply. Asked out loud often enough, it may become a question she can ask herself.
What talk did not do
You might hope that talk like this, built into a school day, would raise her test scores, or make her a better thinker in general. The biggest, most careful trials we found tested the test-score hope, and came back small or empty.
In three trials funded by England’s Education Endowment Foundation, schools were assigned by chance either to try a program or to carry on as usual. One took in 76 primary schools. In half of them, teachers of nine- and ten-year-olds were trained to run structured, reasoned talk for two terms. On progress tests in those subjects, their pupils made about two months’ extra progress in English and science, and one in math, compared with pupils in the other half. Those are small effects. The math result could have been chance, and the English one was borderline. A “month” is the foundation’s way of comparing programs, not growth measured in any child. The evaluators said the results needed caution because of missing data. A year later, on England’s national tests, the same schools showed no clear advantage, possibly because many had stopped using the approach.
The other two tested a program of philosophy discussions, Philosophy for Children. A first trial, with 8- to 10-year-olds in 48 English primary schools, reported about two months’ extra progress in reading and math (on the same yardstick), a small effect, and none in writing. A second trial, larger and more careful, in nearly 200 English primary schools, found no evidence of any effect on reading or math. Nearly half the schools that tried it did not run it as planned. Its authors added that the first trial’s data, analyzed their way, would probably not have shown that gain.
None of these trials measured argument writing.
The better-thinker hope gets mixed results at best. A review that pooled studies of classroom discussion programs, mostly with 9- to 12-year-olds, found that several got children talking far more, and understanding what they read better on some measures. Few measurably improved critical thinking and reasoning. A cognitive scientist who looked at thinking programs argued that reasoning well depends on knowing the subject, and that the gains shrink the further a test moves from what was practiced. That may be part of why the zoo essay was hard. Arguing well about zoos takes knowing something about zoos.
So the test-score hope doesn’t hold up in the largest trials, and the better-thinker hope gets thin support. Your question at the kitchen counter was smaller anyway. It was about one essay, and the paragraph that wasn’t in it.
The class that wrote more essays
In a three-year study at one city middle school in the United States, researchers started with three classes of sixth graders, all 11 or 12 years old. Most were Hispanic or African American, and most qualified for free or reduced-price lunch. Two of the three classes, picked at random (whole classes, not children, so not a randomized trial), met twice a week for a 50-minute class built around argument. On the timetable, it was called a philosophy class. The third class was the comparison. With only one comparison class, the classes could also have differed in ways the program didn’t cause.
In the argument classes, students took sides on social questions. They prepared in small groups of classmates who shared their side, each group with an adult coach, getting ready for a real pair on the other side. Then, pair against pair, they argued it out. Most of those arguments were typed, in online chats, and during those sessions each pair wrote a short reflection on the exchange; in the third year, students went back over their chats in writing, move by move. There was a live, spoken debate as well.
The comparison class was not a neglected one. It had teacher-led, whole-class discussions of the same issues, plus some extra activities and an essay at least every two weeks. Over a year, it wrote more than three times as many essays as the argument classes: about 14 against 4. It did not lack essay practice. It lacked the rest of the program: a classmate on the other side to get ready for and to answer, and written reflection on those exchanges.
Why might arguing with a classmate change an essay at all? The authors borrow an idea from another scholar, Graff. Arguing with a real person, Graff suggested, supplies the “missing interlocutor”: the person on the other side, who gives a written argument its point. That is the authors’ explanation, not something the study tested. Or, to put it our way: an essay with no one in it to answer is an argument with an empty chair. The program had many parts, the getting ready and the written reflection among them, and the authors say it still “requires further experimental dissection” to know which parts mattered.
The test was an essay on a question none of the classes had argued in the program. At the start, the two groups looked alike. In each, about a third of the students made at least one argument in that essay that weighed the other side. By the third year, about four in five of the students in the argument classes did. In the class that had written all those extra essays, it was about three in ten. On a second essay, also on a question nobody had argued, the gap was much the same.
On that second essay, the arguing students’ essays were not measurably longer and held about as many arguments. What changed was what the arguments did: they took up the other side. The arguing students also showed more awareness that evidence matters to an argument.
In later work by the same group, the change came slowly. In one study, evidence use showed up first in the arguments themselves, then in essays on the same topic, and only partly on a new one. It came faster with an added prompt to use evidence. In another, students’ writing took on the other side gradually, on topics they had studied deeply. That fits the zoo essay: it is easier to answer an objection about something you know well. A team from outside the group found that high-school classes that argued about literature in structured discussion improved their argument writing more than classes taught close reading. And arguing is not the only road: studies that teach writing strategies directly also report large gains in essay quality.
The fear behind “don’t talk back” is that letting her argue will only dig her deeper into her own side. In that school, after three years of getting ready with a coach and arguing it out with classmates on the other side, more of the children came to weigh the other side in their essays, not fewer.
The opposite of talking back isn’t silence. It’s a reason.
Look back at Thursday with that in mind. What went wrong was not that she argued. It was that the argument had no reasons in it, on either side, and nobody was answering them. “Everyone else keeps theirs” is not a reason. At 6:40, “because I said so” wasn’t much of one either. That isn’t a failing. It is what tired sounds like.
A good kid is not one who never disagrees. She is one who has learned to disagree with reasons, and to hear the other side’s. That is a skill. In these classrooms it was practiced the way many skills are, by having an argument: not a fight, and not a speech, but a back-and-forth with reasons on both sides and a classmate on the other side who argues back, inside a structured program.
Thursday, a year on
None of these studies ran at a kitchen table, so what follows is a picture of how it could go, not a result.
It is 6:40 on a Thursday, a year on. Pasta again, a little overdone again.
Your daughter is thirteen. She wants to quit the swim team, halfway through the season. She has clearly been planning the ask since the car. It isn’t a house rule. It’s a decision she has brought to the table, and that is different.
“Okay,” you say. “Best reason.”
She has one, and it is a real one: the practices eat every weeknight, and she has stopped liking the water. You give her yours, the real one. She told the coach she would finish the season, and her relay needs four swimmers.
She goes quiet for a second. That is new. Then: “What if I finish the season and just don’t sign up for spring?”
That is not talking back. That is a claim, changed so it can survive an objection. You say yes before you have quite decided to.
Some nights you still say the rule is the rule. The phone still sleeps in the kitchen. She still rolls her eyes. Some rules are not on the table, and she knows which ones. Knowing which ones is part of what changed.
On Sundays there is a question at dinner that nobody has to win. The first one was yours, tried one Sunday last fall, half expecting an eye-roll. You got the eye-roll. She answered anyway. This week it is the zoos again, because she asked for a rematch. The one rule at your table is small: before you argue your side, say the other side’s best reason, well enough that someone who holds it would nod. Taped inside a cupboard door are four questions, in her handwriting now, not yours. What’s your best reason? What would someone who disagrees say? Is that a weak objection or a real one? Can you change your claim so it survives?
Her little brother, who used to agree with whoever spoke last, tries it. His best reason for the other side is that zoos have snacks. She lets it go, mostly. Most weeks it lasts ten minutes. Some weeks it lasts two. Nobody keeps score.
Her essays still come home with green ink. You are hoping, not counting on it, that one of them will someday have a paragraph in the middle that begins, Someone who disagrees would say… At the table, at least, she has started saying it out loud.
You are not running a debate club. On a weeknight you are the person across the table who gives your reasons and asks for hers. On Sundays, with no rule at stake, you are mostly the one keeping the floor fair, so the loudest voice isn’t the only one. It may be most of the job.
If it goes something like this, she will be learning to make an argument the way, in those classrooms, it seems to have been learned. By having one, with someone who answers.