When Is It Time To Say Goodbye?

12:01 minutes

Imagine you’re having a conversation with someone. You may get the sense that they have somewhere else to be. Or you might start feeling restless, and use an excuse to cut the conversation short. Sometimes, you feel like you could talk for HOURS. Chances are you’re wrong every time. 

In a study published this week in the Proceedings of the National Academy of Sciences, Adam Mastroianni and colleagues tried to figure out how good humans are at judging the ideal length of a conversation. They found that both participants agreed a conversation ended at the right time in only 2% of their trials. And the difference between one partner’s desired conversation length and the actual length of a conversation could be as much as 50%—so in a 10 minute conversation, your partner might have wanted to talk to you for as little as 5 minutes, or as much as 15 minutes.

SciFri’s Charles Bergquist talks with Mastroianni about these results, and why the “exit ramps” to a conversation are rarely where you want them to be.

Further Reading

  • Read the full study in the Proceedings of the National Academy of Sciences

Donate To Science Friday

Invest in quality science journalism by making a donation to Science Friday.


Segment Guests

Adam Mastroianni

Adam Mastroianni is a doctoral candidate in the Department of Psychology at Harvard University in Cambridge, Massachusetts.

Segment Transcript

IRA FLATOW: This is Science Friday. I’m Ira Flatow. A little bit later in the hour– everything you wanted to know about baby teeth and an update on the hunt for the Tasmanian Tiger. But first, a conversation about conversations. Here’s Syfy’s Charles Bergquist. Hi, Charles.


IRA FLATOW: All right, so fill us in on this. What’s this all about?

CHARLES BERGQUIST: Well you talk to a lot of people on this show, so I’m going to regard you as the in-house conversational expert.

IRA FLATOW: Oh really? OK, go for it.

CHARLES BERGQUIST: How do you know when a conversation is over, that you’re done talking to somebody?

IRA FLATOW: Well I think that when I’ve asked all the questions and heard all the answers, usually my conversation is over.

CHARLES BERGQUIST: OK, that’s fair, but imagine that it’s not an interview. You’re not going for some specific piece of information here. This is just a random social situation, like you’re chatting with someone at a picnic or a party, if you remember what those are like.

IRA FLATOW: Oh yeah. Well back when we were doing that, if I recall, you would chat until there’s sort of an awkward pause and people are looking at each other. What do I do now? And then somebody will say, excuse me, I need to refill my drink. And then you’re out.

CHARLES BERGQUIST: Right, or take an important phone call or something like that. But have you ever been in the nightmare scenario where you drop that escape line, and the other person just doesn’t pick up on it? They follow you into the kitchen or whatever and they’re still talking.

IRA FLATOW: Oh yeah, I hate when that happens.

CHARLES BERGQUIST: So it turns out, people are universally bad at judging when to end a conversation. Their ideas and the ideas of their conversational partner just aren’t in sync. Adam Mastroianni is a doctoral candidate in psychology at Harvard University. He wrote about this conundrum this week in the journal, “Proceedings of the National Academy of Sciences.” I asked him just how bad we were at this and whether or not it was possible to pin a number on it.

ADAM MASTROIANNI: The number would be 50%. So the difference between what people want and what they get is about 50% of the length of their conversation. Now that doesn’t mean that in all conversations, people would prefer that they were half as long as they were. It means that what people want and what they get differs by half of the length of the conversation, more or less.

CHARLES BERGQUIST: So if I have a 10 minute conversation, my partner might have wanted it to be either 5 minutes or 15 minutes. I like to think of myself as a pretty sensitive person and good at picking up these social cues. But that’s a huge error bar.

ADAM MASTROIANNI: Yeah, well it’s possible that you are the world’s only conversation savant. That is possible, but what we find is when most people try to guess when the other person wants to go, they’re off by even more than half of the length of the conversation. They’re off by more than 60%.

CHARLES BERGQUIST: So did anybody get this right?

ADAM MASTROIANNI: So there’s this sort of unicorn 2% of conversations where both people come out saying, it ended exactly when we wanted it to. If we loosen our restrictions just a bit and ask, in how many conversations does at least one person say it ended when they wanted to? That’s 30%. In 30% of conversations, one person comes out saying, that was good for me. In the remainder, both people say, it didn’t end when I wanted it to. And for most people, it’s pretty far from when they wanted it to.

CHARLES BERGQUIST: Walk me through how you figured this out through the studies that you did.

ADAM MASTROIANNI: Yeah. We ran two studies. In the first study, we got a big sample of Americans and we asked them about their last conversation. We asked them, how long was it? And was there any point at which you felt ready for it to end? If they said yes, we asked, when was that point to the nearest minute? If they said no, we asked how much longer did you want to go?

In our second study we asked those same questions, but we brought people into the lab and had them have a conversation in our lab. And what we found was that most people said, I did feel ready at some point and it was before the conversation ended. But a full 30% of people said, there wasn’t any point when I felt ready for the conversation to end. I wanted more.

CHARLES BERGQUIST: Interesting. So it’s not that we all just get bored with each other sooner than the other person would expect. Sometimes people are skipping out when the other person still is looking for more out of that interaction.

ADAM MASTROIANNI: Exactly. Sometimes both people leave wanting more. One of our most striking results is in the lab. When people have a full 45 minutes to talk– and by the way, plenty of people talk all the way to 45 minutes and we have to cut them off– but even people who leave sooner– 10% of the payers in our second study who spoke in the lab– both left the conversation wanting to continue, and they could have continued.

And we think this is probably because they thought– or at least one person thought– the other person wanted to go. And so they decided to let them go, even though neither of them wanted to go.

CHARLES BERGQUIST: I feel like people are pretty good at picking up certain social cues like, this person is maybe lying to me or, gee, this situation feels somehow unsafe. Why are we so bad at this one?

ADAM MASTROIANNI: That’s funny. I mean, people are actually not that good at picking out liars. They tend to assume that people are telling the truth much more often than they are. But I think of this not so much as people being bad at picking up cues, I think of it as us being very good at hiding the cues.

So when you feel ready for a conversation to end, you generally don’t tell anybody that, at least not until much longer after you’ve felt that way. You might do things like shift around a little bit or break eye contact or maybe not reciprocate a question– all things that are pretty ambiguous. You could be doing those things for a lot of different reasons. Maybe you’re thinking. Maybe you’re distracted. Maybe you really do want to go. But it’s hard for me to know on the other side of the conversation, which I think is why it’s so hard for people to tell when someone else wants to go.

CHARLES BERGQUIST: So it’s not that I’m bad at reading you, it’s that you’re really good at politely fooling me?

ADAM MASTROIANNI: Yes, I think we’re very good at being polite. And polite is an interesting word for it, because I think we think of politeness as a thing that we do with strangers. But we find the same results for strangers, friends, family members, spouses, and lovers in our studies.

And I think it’s because the word politeness means one thing when we’re talking to strangers, but we do the same thing with the people that we love and we call it kindness. But it’s the same thing, fundamentally.

CHARLES BERGQUIST: So is this just a problem of me knowing what the other person cares about in the conversation, or is it that we want two completely different things?

ADAM MASTROIANNI: It’s both of those things. So we find that people rarely want to leave the conversation at the same time, which is what creates the problem in the first place. We can’t both get what we want if we want different things.

Now normally when people want different things, they try to compromise. But in order to compromise, I need to know what you want and you need to know what I want. And so not only do we have this coordination problem, but we have no way of solving it.

CHARLES BERGQUIST: Does it matter who in the conversation is speaking? Like if I’m talking a lot, do I naturally think the conversation is going better than it is?

ADAM MASTROIANNI: It’s a complicated question. I don’t know if we know the answer. We do know, at least, that it doesn’t really matter who ends the conversation. And so maybe by extension, it doesn’t really matter who is doing more of the talking in the conversation. When we ask people afterward, who took the first step toward ending the conversation? In 90% of pairs, people agree on who that was. And that wasn’t a given. It could have been that everybody claims that they were the prime mover.

But in general, people agree on who it was who started the closing ceremonies. But when we look at whether it makes a difference– whether you were that person or whether you were on the receiving end of the end, it doesn’t make any difference. It’s as if we all saw the exit on the highway and one person said, we should get off there. The other person said, we sure should. And that’s where it ended. Unfortunately that exit wasn’t exactly where either of us wanted to exit the highway.

CHARLES BERGQUIST: So those conversational exit ramps are things like, oh, would you look at the time? Or I really need to go check on people in the kitchen and see what they’re doing, or that sort of thing.

ADAM MASTROIANNI: Exactly, yeah. So the thing about a conversation is that it is a lot like driving down the highway. You can’t just leave whenever you want to, at least not without causing a lot of damage. You’re going to run into a tree or a storefront or into a ditch. And conversations are similar.

There are pretty strong rules about when you are allowed to leave. You can’t leave in the middle of someone’s sentence, unless there’s some kind of emergency. You can’t leave in the middle of a story. And even broader than that, you’re supposed to reciprocate questions. And you could really just feel that most of the time in a conversation, it’s not an appropriate time to go, all right, see you later.

We think there’s a lot of distance between those exit ramps, and that’s why we get off on them sooner than we want to or later than we want to. Just like when you’re on the highway, you wish you can get off exactly when you want to and go straight to your destination. But often, you have to wait a couple miles or get off a couple of miles early.

CHARLES BERGQUIST: I’m thinking about good friends who have known each other for a long time and they’re together. They may chat for a bit and then there’s just that long, companionable silence while they’re doing something, and then they start talking again. But then when we’re dealing with somebody that we don’t know, that’s not a thing. That would not be natural to be silent with somebody that you don’t know in a conversational setting. Do we know anything about that?

ADAM MASTROIANNI: I don’t know if we know anything about the silences that arise in conversations between friends versus between strangers. My intuition is that the reason you can have a silence with a friend is because it’s not really in question whether you like them or they like you. Whereas with a stranger, you don’t necessarily know. A silence might mean that they’re ready to be done talking to you. And you might know that that’s not the case with a friend.

I think a good thought experiment is talking to a stranger on a train, where you know that if you stop talking, it’s not because you can walk away. And so you might drift in and out of conversation. And it doesn’t necessarily say anything about how much each party likes the other, it’s just more about the constraints of the situation.

CHARLES BERGQUIST: Right. We’ve all been spending a lot more time interacting online these days. How does that play into this? Is there any difference in online communications versus in-person communications?

ADAM MASTROIANNI: We don’t know for sure if this is different online. We ran all of our studies ran pre-pandemic. Now the idea of talking to somebody face-to-face sounds both quaint and desirable. But my guess is that it would be [INAUDIBLE] if we weren’t in person, for a few reasons.

One is that even though we’re not great, necessarily, at picking up the queues, I think mainly because we’re good at hiding the queues, there are even fewer queues. You don’t get to see people’s eye contact as well. You don’t see their body language.

But maybe more importantly, there’s no plausible reason to end the conversation when we’re online. If I were to tell you that I’m ready to be done talking to you, what am I going to go do, stand alone in my kitchen? There is nothing really going on, so I don’t have any excuse to get out of this conversation.

CHARLES BERGQUIST: Are the people that you would say– this person is a great communicator– necessarily better at this conversational judgment than other people?

ADAM MASTROIANNI: We don’t know for sure. There’s some research from some of our colleagues at Cornell, where they asked a whole dorm-full of people to nominate who the best and worst conversationalists were. What they’ve told me is that pretty much everybody gets nominated as both the best and the worst, which suggests that there’s maybe not such a thing as a great conversationalist. There’s a great conversationalist for you. And I think the mark of a great conversationalist is someone that makes you feel liked, that makes you feel good, that makes you want to talk to them. And that might be a different person for each person.

CHARLES BERGQUIST: Do you have any tips for how we can become better at this?

ADAM MASTROIANNI: I don’t necessarily have a tip for how people can become better perceivers. I do have a tip for how people can enter conversations more gracefully. And I think it’s by addressing the problem of partying head on.

So the reason the ends of conversations are so fraught is because it feels inherent to the parting of our ways that something has gone wrong, otherwise we’d still be talking to each other. If we really liked it, why don’t we just keep doing it? And so I think a lot of the energy that we put into ends of conversations are trying to overcome that problem.

And I think the best way to do it is to address it head-on by saying, I had a really nice time talking to you. I’m looking forward to talking to you again. You don’t have to lie. You don’t have to say, I’ve got to be somewhere.

But just address the fact that this conversation isn’t ending because anything went wrong. It’s just that sometimes, in the course of human events, two people have to stop talking to each other and do something else. And this is one of those times, but I’m glad for the time that we had together.

CHARLES BERGQUIST: I’m going to cheat and look at the clock and that tells me that it is time to end this conversation. Do you agree?


ADAM MASTROIANNI: I think this conversation is ending exactly when I wanted it to. It’s been lovely talking to you.

CHARLES BERGQUIST: Adam Mastroianni is a PhD candidate in psychology at Harvard University in Cambridge, Massachusetts For Science Friday, I’m Charles Bergquist.

Copyright © 2021 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of Science Friday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our policies pages at http://www.sciencefriday.com/about/policies/

Meet the Producer

About Charles Bergquist

As Science Friday’s director and senior producer, Charles Bergquist channels the chaos of a live production studio into something sounding like a radio program. Favorite topics include planetary sciences, chemistry, materials, and shiny things with blinking lights.

Explore More