How Do Rumors Get Started? How do rumors spread from person to person? And why do some rumors spread like wildfire, while others never seem to make it out of a small circle of interconnected people?

How Do Rumors Get Started?

  • Download
  • <iframe src="https://www.npr.org/player/embed/6429833/6429834" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

JOE PALCA, host:

All right. You're standing around the water cooler and one of your co-workers says, You know what I heard? Or your mom calls, or maybe she e-mails you. That's the new way to have your mothers watch over you and to tell you something that's dangerous that she's heard from one of her friends. It's how rumors get started. But is there any rhyme or reason to how rumors are transmitted or what makes for a good rumor and why some rumors spread like wildfire and others never make it out of a small group of people?

Well, there may be answers to these questions, and joining me now are two guests who can help provide them. First, Nicholas DiFonzo is the co-author of the new academic book Rumor Psychology: Social and Organizational Approaches, just out from the American Psychological Association. He's a professor of psychology at the Rochester Institute of Technology in Rochester, New York. And he joins me from the studios of member station WXXI in Rochester. Welcome.

Professor NICHOLAS DIFONZO (Rochester Institute of Technology): Hi, Joe. Glad to be here.

PALCA: And also with us this hour is Duncan Watts. He's the author of Six Degrees: The Science of a Connected Age, first published in 2003. He's a professor in the Department of Sociology and the Institute for Social and Economic Research and Policy at Columbia University in New York. And he joins us from our New York studios in New York. Welcome back to SCIENCE FRIDAY, Dr. Watts.

Professor DUNCAN WATTS (Columbia University): Hi, Joe.

PALCA: And we'll be taking your calls, so ask us your questions or maybe tell us your rumors: 1-800-989-8255. 1-800-989-TALK. And I'm afraid we're going to have to go to a break in just a few - a minute or so, but I thought we could at least get started here with a question for you, Dr. DiFonzo. What is the difference between, let's say, rumor and gossip or rumor and urban legend?

Prof. DIFONZO: Well, rumor is unverified information and circulation among a group of people who are usually trying to make sense of an unclear situation or circumstance or they're trying to manage a threat or a potential threat. So the classic example of a rumor would be, hey, I heard that NPR was downsizing, Joe. What did you hear?

PALCA: Uh-oh.

Prof. DIFONZO: So people would, in this situation...

PALCA: I heard that too.

Prof. DIFONZO: I hope it's not true. The people in this situation are trying to figure out the facts. They're trying to make sense of the situation so that they can diminish a threat or potential threat that's coming their way. Whereas gossip has much more to do with social chat - social talk - that is ostensibly less important or less outcome-relevant that rumor is, and it's all about connections. So if I share with you gossip - did you hear what Duncan did at the party the other night...

PALCA: Well, we're going to have to hear the end of that juicy tidbit after we come back from a break, so stay with us. I'm sorry. We will be back in just a short while talking about rumor and gossip and all that good stuff.

(Soundbite of music)

PALCA: From NPR News this is TALK OF THE NATION: SCIENCE FRIDAY. I'm Joe Palca. We're talking about rumors, gossip and urban legend, and who doesn't like talking about rumors, anyway?

My guests are Duncan Watts. He's a professor of sociology at Columbia University at New York, and Nicholas DiFonzo, professor of psychology at the Rochester Institute of Technology in Rochester, New York.

And we'd like to hear from you. Our number is 800-989-8255. That's 800-989-TALK. And the rumor that we're changing that number is completely untrue. It will be the correct number for the entire hour. And just before we left on a break we were talking with Dr. DiFonzo about the difference between rumor and gossip. And rumor is something that may be threatening that you may or may not want to do something about, but gossip is a little different, and maybe you could finish that explanation.

Prof. DIFONZO: Sure. Gossip is all about social chat. When I share gossip with you, you're my friend. I feel closer to you and you feel closer to me, so there's a bonding function that goes on. And I could also be talking about things that the group ought to be doing or ought not to be doing. So if I talk about Duncan's escapades at the party last Friday night, I could be criticizing him for doing that, and by the same token, talking about what people should not be doing at the party, or if I'm lauding him, what they should be doing at such occasions.

So gossip is very social in nature. It's all about the personal social networks that one finds one a part of.

Urban legend is a narrative story. It has a plot, a setting, a climax, usually a denouement, and almost always a moral to the story. So they're usually intricate stories about modern themes such as hitchhikers and automobiles - those sorts of things. They're entertaining. There's some bonding that goes on as well, but they usually make sense of the world in a broader context. That is, they help us figure out the world and they have a moral to them much like the old fables did.

PALCA: Hmm. Let me turn to you, Duncan Watts, and ask the question - it's always struck me that it sometimes feels that there's one person somewhere in the world who's generating all these things, and if you could just figure out who that person was and tell them to stop, it wouldn't happen anymore. But I guess it couldn't be that way. But is it ever possible to say that rumor started with him or her?

Prof. WATTS: In any particular instance, it's obviously, you know - everything starts somewhere. You know, if you make the analogy with epidemics of infectious disease, you know, even sort of global pandemics like the 1919 Spanish flu, you know, almost certainly started with a single person somewhere, probably, you know, in Southern China. And the same thing is going to be true of rumors, whether they spread just very narrowly within a single group or spread very widely.

The point to make, though, is that there needn't be anything particularly special about the person who starts the rumor, and so this is sort of a bit counterintuitive, because when something special happens - you know, when some, you know, amazing rumor starts sort of sweeping across a city or a country or even throughout an organization, you might think that the person who began it had to be special as well. But that actually turns out not to be the case necessarily.

You know, it has - you know, whether - why some things spread and other things don't, you know, is partly, you know, due to the kind of psychological affects that Nicholas is talking about. You know, some things just so to satisfy our emotional needs than others, and that's certainly relevant.

But there's also a structural side to the problem as well, and this is sort of more where the social networks become relevant, is that people are not, you know, uniformly spread across, you know, geography or through cities or through organizations. They are clumped together. They are connected non-randomly in all sorts of sort of complicated and overlapping networks, and the structure of the groups and the organizations and the way that they're connected to each other can have, you know, enormous impact on what spreads and what doesn't, and that's really not something that you can pin on any particular individual.

PALCA: Okay. Well, let's invite our listeners to join in this conversation. Our number is 800-989-8255. And let's first go to Ian in San Diego, California. Ian, welcome to SCIENCE FRIDAY.

IAN (Caller): Hi, thanks for taking my call. This is a great program. I'd like to propose that social - that viral marketing is sort of a reinvention of the rumor for profitable purposes, specifically in the online world, but I'd like to hear your guests' comments on it.

PALCA: Okay. Well, Dr. DiFonzo, are you familiar with the concept of viral marketing, first of all?

Prof. DIFONZO: Sure, viral marketing. It uses the epidemic kind of metaphor or analogy that a rumor spreads by word of mouth through these different networks that Duncan and others are studying. And it's cheap. I don't have to pay for airtime in order for things to spread far and wide, and it's often very effective. Who's the best person to persuade me but somebody that I know, somebody in my personal social network, that Product A or Product B is good?

PALCA: But how does the virus, in this case, or the rumor, get injected into the system? I mean, where does that germ come from? Go ahead, Dr. Watts.

Prof. WATTS: Well, you know, this is sort of - there's a bunch of different strategies, of course. You know, you can - you can start with some sort of online campaign and you can try to get people talking about some mysterious video or some sort of funny website and give them tools so that they can easily forward URLs or e-mails to their friends, for example, and that's how, you know, some of these viral campaigns get seeded. Other times you can, you know, you can hand out free merchandise at parties or at bars or, you know, leave them with random people in their mailboxes.

The problem is that nobody knows yet how to sort of engineer - to make use of these - these, you know - clearly in some sense these word-of-mouth ties are very useful and people trust them and pay attention to them in ways that they don't pay attention to traditional media. The downside is that it gets very complicated and it's hard to sort of engineer these networks to do the things that you want them to do.

PALCA: All right. Ian, thanks very much for that question. Why don't we go now to Marcus. I'm sorry, Marcus in Virginia. Welcome to the program.

MARCUS (Caller): Hi. I'm working for an Internet company and we've got layoffs pending, and the only source of information that we really have is rumors, and they tend to be a lot more reliable than official channels, and I'm just wondering if you had any comments on rumors - you know, the growth of rumors when either official channels aren't reliable or there just isn't much information.

PALCA: Yeah, Dr. DiFonzo, what about that? Do rumors ever - you know, I mean, is there usually a grain of truth somewhere, or is that not always the case?

Prof. DIFONZO: Marcus is in a very common situation, in which rumors flourish when there's a lack of formal information. The best way to get rumors going in an organization is to not saying anything or not say very much or to say contradictory things. And so it's not surprising that the rumors are spreading very quickly in that organization.

The other thing to say is that the most interesting question to me about rumors is, how do you know that what you know is true? That is, how do you know that this rumor is either accurate or not accurate? And it turns out that there's some situations in which they are very accurate, and there's other situation which they're not very accurate. And looking at a collection of communications studies that were done in the '70s, '80s, where communication researchers went into organizations such as the one Marcus is in and they collected rumors and they calculated the percentage of rumors that were accurate.

Turns out that in organizations where there's an established grapevine, the rumors are usually 95 or more percent correct. And so, Marcus, you're doing well. I would bet money on some of those rumors that you're hearing, especially if the network that you're in is an active one and the people are interested in figuring out the facts rather than fulfilling some kind of prejudice or bias, and if they're - and if it's an established grapevine that's been around for a while.

So if I hear something from my co-worker, I've known him for a while and I can judge his credibility.

PALCA: Okay...

MARCUS: I...

PALCA: Go ahead. Marcus, go ahead.

MARCUS: Generally what I try and do is just crosscheck everything.

PALCA: Yeah.

MARCUS: What I hear from one person I try and see if I hear it from others.

PALCA: Yeah, yeah. Well, that's, you know - I mean that's - everybody's singing from the same book, but it might be from the wrong one, you know?

MARCUS: That's happened too.

PALCA: Marcus, thanks very much for the call. You know, not very long ago, just speaking of social networks and people liking to check things, there was a rumor flying around NPR that we were about to have a shakeup in senior management. And you know, this is a group of reporters here. So they're sort of used to tracking things down. And sure enough, the next morning the shakeup was announced. So somebody knew something and it was accurate. A funny position to be in, though, when it's in your own newsroom.

All right. Well, let's take another call now. Rebecca in Des Moines, Iowa. Welcome to the program.

REBECCA (Caller): Hi.

PALCA: Hi.

REBECCA: I don't know if it's an urban legend or a rumor - certainly it's in the political arena - but the biggest rumor that I've heard, or urban legend that I've heard, is that Al Gore claimed to have invented the Internet. He never said that he invented the Internet, but the more the story got told, the more people believed that it was true or believed that it was fact, and it was never the case.

So I don't know, I guess I would like your comment on that particular urban myth or rumor.

PALCA: Okay. Thank you for the question, Rebecca. What about - Duncan Watts, first of all, how would you characterize that, urban legend or rumor?

Prof. WATTS: Well, I guess I am sort of - I guess I don't have the same taxonomy of how things spread. I tend to sort of, you know, think about these problems from something of a social learning perspective that, you know, basically the world is very complicated and it's impossible for all of us to go out and gather all the information that we need on a daily basis.

And so it makes perfect sense to try and infer information from other people. So we have some sort sensible heuristics for saying, well, you know, if I hear something once, then that increases the likelihood that I'll believe it. And if I hear it twice or three times then, you know, that likelihood increases in some possibly sort of non-linear kind of way.

And so, you know, one of the important and possibly the most important property of a rumor is whether you believe it and transmit it the first time you hear it or whether you have to hear it twice. And just that one step makes an enormous difference in how far things spread and how fast they spread.

So possibly in the Al Gore case, you know, it's not true, but it's sort of, you know, it's kind of easy to believe and it's sort of fun to talk about. So you have a relatively low threshold for accepting it as the truth. And those are the things that tend to get around very widely.

PALCA: Dr. DiFonzo, I'm just wondering if there's any way to predict in advance how people try to inject in rumors and then make some sort of calculation about whether they're going to spread or not.

Prof. DIFONZO: There were several studies some time ago where researchers would go into an organization or they'd go into a group and they would inject a false rumor, usually pretty innocuous. And then they would wait a few days and then pass out questionnaires to everybody in the organization and say have you heard this rumor and where did you hear it from, and they'd trace it back to the person who spread the rumor.

Now, the Al Gore rumor is - we could kind of classify that also as an urban legend. There is some overlap sometime. But I think the interesting thing about that rumor is that - it is that Al Gore said he invented the Internet. And if you think about it, the motivation for spreading that kind of rumor is to kind of make fun of Al Gore. So obviously the people who would be most likely to spread that rumor would at least be in part motivated by a negative disposition toward Al Gore and that sort of thing.

PALCA: Got it.

Prof. WATTS: That's my take on that rumor for Rebecca.

PALCA: Okay. All right. Well, thank you for that. And we're talking about rumors, gossip, urban legends, how they spread. I'm Joe Palca and this TALK OF THE NATION from NPR News.

And let's take another call now. And let's go to, oh, why don't we say Bob in Wilmington, Delaware. Welcome to the program.

BOB (Caller): Thank you very much. I take a lot of pleasure in debunking urban myths, especially at work, like using the Snopes Web site. But there's one thing I never quite understood. Prior to the Internet, back in the - especially back in '80s, I've never quite understood how urban legends could spread so quickly back then, you know, in sort of a pre-technological age. I'm thinking of one story in particular I've heard. I don't want to go into great detail. But it's a local Philadelphia TV personality, something sexual about this person that spread incredibly rapidly. And anybody I know who grew up in that area knows this story. And it seems like overnight everybody knew this particular story about this guy.

PALCA: They're raising their hands in the control room here saying we know that story. I don't know the story. But yes, please don't go into any details here...

BOB: I'm not going to. How did that, pre-Internet, how did something like that circulate?

PALCA: I know. It's hard to believe we did anything pre-Internet. But okay, let's go to Duncan Watts. Maybe he can help us out a little bit here.

Prof. WATTS: So I think there actually is a good reason for that. And it goes back to this sort of well-known idea of the small world problem that social networks have a certain structure to them such that, you know, everybody is connected in a sense by only a few degrees of separation. So you know someone who knows someone who knows someone who knows anyone in the world. And that is, you know, an idea that's been around for a long time.

But I think that recently we've gotten to understand it very well. And one of the things that we've understood about it is that this small world probably happened quite a long time go. And so, you know, and once the, and the reason is that you don't require a lot of fancy technology for social networks to develop this kind of property.

And the second thing is that once the world gets small, it's much harder to make it smaller, right? So in some sense, yes, the world is getting smaller and these rumors probably are spreading a little bit faster as a result of the Internet and e-mail and so on.

However, they were already spreading very fast. And so you don't get a, you know, you're sort of incurring diminishing returns at this point.

PALCA: Okay. Bob, thanks very much for that question. Dr. DiFonzo, we only have about 30 seconds left. But I wonder if I can ask you briefly, is there any way, once a rumor gets started, to stop it?

Prof. DIFONZO: Well, there is some good techniques that one can use if you're in that position where you want to stop the rumor. First of all, I'd urge you to stop the rumor only if it's not true. You wouldn't want to deny a rumor or rebut a rumor if in fact it is true. Don't use the approach no comment, or rather almost never use the approach no comment, because it increases a level of uncertainty: why is this person saying no comment?

One of the best techniques is to rebut the rumor, but to use a third party, a trusted third party to do the rebutting. And then when you rebut the rumor, to provide a context.

PALCA: Well, we'll have to leave it there, I'm afraid. Thank you very much both of you. Dr. Nicholas DiFonzo is a professor of psychology at Rochester Institute of Technology, and Duncan Watts is the author of Six Degrees: The Science of a Connected Age.

Copyright © 2006 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.