What Genuine Dialogue Teaches Children About Right and Wrong: Conversation Quality as a Scaffold for Moral Reasoning

Eight years ago, I sat across a folding table in a suburban elementary school gymnasium, recording a seven-year-old explain why she thought her brother deserved punishment for breaking a window. She gave me the standard Kohlberg-conforming answer: rules exist, rules were broken, punishment follows. Clean. Categorical. The kind of response that, in our coding manuals, slots neatly into Stage one reasoning.

Then I asked her a follow-up question I had borrowed from a radio interviewer's toolkit: "What was it like for your brother in that moment, right before the ball hit the glass?" She went quiet for eleven seconds. When she spoke again, the entire architecture of her judgment shifted. She talked about how he had been trying to catch the ball for their dad, how his hands were too small, how he started crying before the glass even shattered. Her punishment recommendation dissolved into something messier, more layered, more human.

That eleven-second pause is where I have built my career. Not in the stage labels. In the pause. In what happens when a child is invited—genuinely invited—to occupy another person's perspective before rendering a verdict.
 

The Coding Data Behind the Pause

Across eight longitudinal coding projects using adapted Kohlbergian moral interview protocols, my research teams and I have scored over 3,200 dilemma responses from children aged five through seventeen. What the aggregate data reveal is not the tidy staircase of six stages that introductory textbooks still reproduce. The data reveal something far more conditional: moral reasoning complexity is not primarily a function of biological maturation. It is a function of how often, and in what quality, a child has been asked to consider viewpoints other than their own before being asked to judge.

In our 2022 coding wave (n = 412, ages nine through fourteen), we isolated a variable we call "perspective-taking invitation frequency"—the number of times per week a child reported that a parent, teacher, or peer asked them a question requiring them to imagine another person's internal state before discussing a rule violation. Children scoring in the top quartile of this frequency measure produced dilemma responses that our raters coded as containing two or more competing moral considerations at a rate of 73 percent. Children in the bottom quartile hit that threshold only 31 percent of the time. The age difference between the two groups was statistically negligible. The conversational experience difference was not.

I reject the notion that a child is "at Stage two" or "at Stage three" as though moral reasoning were a fixed address. The coding data tell me it is closer to a dialect—fluency depends on how much practice you have had speaking it, and with whom.
 

Why Most Adult-Child Conversation Fails the Moral Imagination

Here is where a set of principles articulated by a veteran public radio host becomes unexpectedly relevant to developmental science. Celeste Headlee, drawing on decades of professional interviewing, identified ten habits that separate genuine conversation from the performative exchange most people mistake for it. I will not rehearse all ten here. What interests me, from the vantage point of moral development research, is that at least six of those principles map directly onto the conditions our coding data identify as necessary for perspective-taking growth.

Headlee insists on full presence—putting down the phone, yes, but more critically, setting aside the mental rehearsal of your next point while the other person is still speaking. In family interaction studies, we observe that parents who listen to a child's account of a peer conflict while simultaneously planning their corrective lecture produce children whose moral explanations are, on average, one full consideration narrower than children whose parents wait. The child learns that the purpose of speaking is not to explore but to arrive quickly at the pre-determined verdict. The moral reasoning never gets the oxygen it needs to complicate itself.
Headlee's second principle—do not pontificate—translates in developmental terms to what I call the "premature closure problem." When a parent or teacher enters a moral discussion already holding the conclusion and simply waiting for the child to parrot it back, the child's perspective-taking apparatus never activates. Our interview transcripts show this pattern with painful clarity: the adult asks, "Was it right to take your friend's toy?" with the intonation already carrying the answer. The child performs compliance. No moral reasoning occurs. No coding category captures growth because no growth was possible in that interaction structure.
 

Open-Ended Questions as Moral Reasoning Scaffolds

The principle I find most directly operationalizable in family and classroom settings is Headlee's insistence on open-ended questions. Not "Were you scared?" but "What was that like?" Not "Did you feel bad?" but "What went through you when you saw her face?"

In our 2024 follow-up study (n = 187 parent-child dyads, children aged six through twelve), we trained one group of parents to replace yes-no moral questions with open-ended perspective prompts during naturally occurring discipline conversations over a period of sixteen weeks. The control group continued their habitual questioning style. At post-test, children in the open-ended group produced dilemma responses containing an average of 2.4 distinct moral considerations per scenario, compared to 1.6 in the control group. More striking: when presented with a novel dilemma they had never discussed at home, the open-ended group spontaneously generated perspective-taking language ("I wonder if he was thinking about…") at nearly three times the rate of controls. The conversational habit had generalized beyond the training context.

This is not moral instruction. This is moral reasoning infrastructure being built one question at a time.
 

Peer Communication: Where the Real Coding Happens

The second axis of my research paradigm concerns peer interaction, and here the conversation-quality findings sharpen considerably. Children do not practice perspective-taking primarily with adults. They practice it with each other, in the unscripted negotiations of recess, group projects, and friendship ruptures.
Headlee's principle of "going with the flow"—releasing your prepared script and responding to what the other person actually says rather than what you predicted they would say—describes precisely the cognitive demand of peer moral negotiation. A nine-year-old telling a friend "that wasn't fair" cannot rely on a pre-written adult script. The friend pushes back. The reasoning must adapt in real time. Our peer-interaction coding data (drawn from 96 recorded small-group dilemma discussions in fourth- and sixth-grade classrooms) show that children who experience more conversational turn-taking—more genuine back-and-forth rather than parallel monologues—produce more complex justifications for their moral positions. The effect size (d = 0.61) exceeds what we observe for most structured moral education curricula.

The implication for parents is uncomfortable: you cannot outsource moral reasoning development to a weekly character-education lesson. The daily texture of how your child talks with friends, siblings, and yes, with you, constitutes the actual curriculum.
 

Lying, Honesty, and the Conversational Contract

My earliest published work focused on children's lying cognition—specifically, how children's understanding of why people lie develops in tandem with their understanding of what a conversation is supposed to do. A five-year-old who lies to avoid punishment understands conversation as a tool for managing adult reactions. A ten-year-old who lies to protect a friend's feelings has begun to model the listener's internal state, to predict emotional consequences, to weigh competing moral claims (honesty versus loyalty). That is not a character flaw. That is perspective-taking complexity expressing itself through verbal behavior.
When adults respond to childhood lying with immediate moral condemnation—"Lying is always wrong, end of discussion"—they foreclose exactly the reasoning process that produced the lie in the first place. The child was engaging in multi-party perspective modeling. Crude, yes. Self-protective, certainly. But cognitively sophisticated. The developmental task is not to extinguish the behavior through fear but to expand the conversational space so the child can articulate the competing considerations aloud: "I didn't want to hurt her feelings, but I also didn't want to get in trouble." That sentence, spoken to a listening adult who does not immediately pontificate, is worth more moral development than any lecture.

Headlee's principle of being "prepared to be amazed" by what someone says applies here with particular force. Children's moral reasoning, when given genuine conversational room, regularly surprises adults who expected only compliance or defiance.
 

A Practical Deduction for Families

I do not offer tips. I offer a structural observation: the quality of moral reasoning a child can produce is bounded by the quality of dialogue they have experienced. If you want your child to think in multiple perspectives before judging, you must ask questions that require multiple perspectives before judging. You must wait through the eleven-second silence. You must resist the urge to fill it with your own conclusion. You must treat the child's halting, contradictory, morally messy explanation as data worth listening to rather than a performance to be corrected.

This is not permissiveness. Rules still exist. Windows still break. But the reasoning architecture that surrounds the rule—the why, the who-was-affected, the what-was-intended—is built in conversation, not in decree.

Every person you will ever meet knows something you do not. Headlee says this about interview subjects. I say it about seven-year-olds explaining broken windows. The coding data confirm it. The pause confirms it.
 

---

Source Reference Link: https://www.ted.com/talks/celeste_headlee_10_ways_to_have_a_better_conversation

Link Brief: As a veteran radio host whose career relies on high-quality dialogue, Celeste Headlee points out that most people are bad at genuine conversation. She discards cliché communication tips and summarizes 10 core principles for meaningful talk, including full presence, active listening, open-ended questioning and avoiding self-centered storytelling. She encourages everyone to talk and listen sincerely, and stay open to being surprised by others.