Judith Donath on Signaling, Design, and the Social Machine
Oct 31 2016

Social%20Machine.jpg Judith Donath, author of The Social Machine, talks with EconTalk host Russ Roberts about the ideas in her book--an examination of signaling, online identity, and online community. Donath argues that design elements in technology play a key role in our interactions with one another. The conversation closes with a discussion of data collection by corporations and the government.

Adam D'Angelo on Knowledge, Experimentation, and Quora
Adam D'Angelo, CEO of the question and answer website, Quora, talks with EconTalk host Russ Roberts about the history, evolution, and challenges of Quora. Along the way they discuss the aggregation of knowledge and the power of experiments for improving...
Kevin Kelly on the Inevitable
Futurist, author, and visionary Kevin Kelly talks with EconTalk host Russ Roberts about his latest book, The Inevitable, Kelly's look at what the future might be like and the role of the human experience in a world increasingly filled with...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Matěj Cepl
Oct 31 2016 at 9:38am

Concerning face-to-face interviews, if I remember correctly Terry Gross (whatever you think about her political views, she knows a thing or two about interviewing) makes Fresh Air interviews by sitting in the other studio than the interviewee, just to avoid nonverbal communication.

Tim Bugge
Oct 31 2016 at 3:12pm

Enjoyers of this interview might also enjoy “Black Mirror” s3 e1 “Nosedive” on Netflix about a world in which one’s social rating via social media becomes the main criteria for the ability to rent homes, get loans, etc.

Marilyne Tolle
Nov 1 2016 at 6:03pm

Very interesting programme but a real shame that the conversation on online signalling didn’t include a discussion of avatars in virtual-reality worlds like Second Life.

The choice of gender, age, height, skin colour, even species (one can be a human or an animal) for the avatar (not even to mention clothing, shoes, accessories and hairstyle) provides a goldmine for the study of “raw” signalling.

Judith Donath
Nov 2 2016 at 1:05am

Hi Marilyne,

Signaling theory looks at 2 categories of signals — inherently reliable ones, where the form and meaning of the signal are linked (and thus having the indicated quality helps or is necessary to produce the signal), and “conventional signals”, where the signal form refers to the mean ony by convention, and thus is not inherently reliable.

Avatar forms are conventional signals. In many avatar worlds it’s just as easy to be a talking bipedal frog,or a princess, or a punk skateboarder – just change your avatar. While this expresses something about one’s taste, the ease of assuming these roles quickly makes them somewhat meaningless – they’re removable masks, not identities.

To me, a key question is: how to develop avatars that have a tighter, more meaningful connection between some underlying trait, and the avatar’s appearance. What traits do we care about? What would reliably signal them?

As I mentioned, the rapid changing of fashion is itself a representation of how information moves with a community – and thus avatars that had fashions would signal their place in that hierarchy. Avatars whose appearance accumulated history (data portraits) would also be more reliable signals.

some references: Diane Crane on fashion, e.g.
Crane, Diana. 2000. Fashion and Its Social Agendas: Class, Gender, and Identity in Clothing. Chicago: University of Chicago Press.
; I have some papers on Data Portraits;
Donath, J., A. Dragulescu, A. Zinman, F. Viégas, and R. Xiong. 2010. Data portraits. Leonardo 43, no. 4: 375-383.
Jacobson, David. 1999. Impression Formation in Cyberspace: Online Expectations and Offline Experiences in Text-Based Virtual Communities. Journal of Computer-Mediated Communication 5, no. 1.
Donath, Judith. 2007. Virtually trustworthy. Science 317: 53-54.

Nov 2 2016 at 9:56am

@ 11:30 Russ Roberts said, “And of course we’re always trying to make an impression in some abstract, subconscious sense. But what I find amazing on Twitter–and I guess, sometimes even in real life, is someone will accuse me of something–being the pawn of some special interest, hating some particular group. And I want to just yell, ‘You don’t know me!’ And of course, I know that already. Why do I need–why do I have an urge, why would I think for a minute that, despite all that I have written online, written in print, said over the year in this program–why should I be surprised that people don’t know me? But it hurts. There’s something–and then I watch other people, when they are insulted, say, on Twitter–and they respond with this virulent vulgarity, say, in response to an insult. And I always wonder, drop them a quiet note and say, ‘Chill out. Don’t feed the troll.’ But then I notice my urge to do the same.”

  • This fits nicely with F.A. Hayek’s distinction between instinctive moral rules—which cannot support a market system—versus evolved morality—which is required to maintain a market system. When reputation or ideas are threatened, instinct tells us that a threat has been made against our person and that such a threat warrants a strong response up to and including physical violence. Evolved morality, on the other hand, abhors violence. Evolved morality [evolved meaning, ‘selected through an evolutionary process’] tell us that words are wind; that verbal slights can–possibly even must–be tolerated and ignored; that insults to our reputation are not equivalent to physical attacks and do not warrant a violent response. This is an improvement because, in the absence of physical aggression, voluntary trades are possible. Voluntary trades improve the standard-of-living of the traders. The fact that Russ Roberts is wealthy and rarely feeds trolls is not coincidence.

At 00:33 Judith Donath said, “Instead we rely on signals, which are perceivable indicators of these otherwise unperceivable traits. However, for the person signaling, it can be profitable to deceive: to say that they are nicer, better, faster, smarter than they actually are. And this–the tension between these two, because being deceived is harmful–this tension has brought about, in the world of evolution an arms race in communication that has shaped a lot of the animal world.”

  • I wonder if Judith meant “arms race” in the literal sense. I wonder if she was making a causal argument that advances in communication are potent determinants of natural selection. If that was her argument, it would find a close advocate in Hayek’s chain of arguments that moral rules set limits on the complex orders people are capable of forming; that the complex orders serve as mechanism for gathering, processing, and communicating information; that much information is communicated through signals—especially price signals; and that improving a system’s ability to gather, process, or communicate information gives a survival advantage to the people who participate in the improved order.

At 44:17 Judith Donath said, “Because we really do live in a world in which access to information is one of the key markers of status. I mean, wealth is certainly a very important one; but the whole notion of how do you get access to information, how good are you at assessing information, is a huge amount of how we display who we are. And that’s really what fashion ultimately is.”

  • She kind of rocked my world with this–elevating what I had up till now considered trivial to the level of profound.
  • What an exceptional podcast.
Deltango Vale
Nov 2 2016 at 7:32pm

Having lived in Second Life for 10 years, I assure you that presentation is everything. We are 100% self-selecting projections, choosing not only from which designer to buy our earrings or which walking, flying, standing, sitting animation to use on a daily basis, but also our hair, eyes, skin, shape and the intimate details for our anatomies.

We judge each other and are judged more stringently on appearance, deportment and style in SL than RL exactly because we are CAN choose 100% of what we wish to project. There are no physical defaults to hide behind. If someone chooses to be tall or short, plain or pretty, blond or redhead, human or furry, everyone asks, why did she choose those attributes? What is she transmitting to us about herself?

I, for example, spent my first year honing my appearance and have stuck with it for the next nine years (I still wear my original hair and eyes). Nearly everyone I know in SL has the same philosophy: to establish a persona, an identity based on appearance and deportment and to maintain it through time. Those who wish an alternative identity create an alt rather then mess with their primary identity.

It is a mistake often made by neophytes to assume that an avatar is a crude cardboard cutout to be used and disposed of at will, but this reflects their ignorance of the sophistication of the world they have entered. If they stick it out for a few weeks, they soon realize that they are at the bottom of the social ladder and must up their game if they are to be taken seriously by the veterans. A free, poorly-made blouse won’t cut it in a popular club; it really matters where you buy your shoes. I have 30+ dance animations, for example, that I switch between rapidly depending on what song is being played by the live DJ. My social status at the club depends on how I present myself. If you think women are competitive in RL, they are 100x more so in SL.

Our inworld identities are reinforced through social media (various fora, SL’s version of Twitter, FB avatar accounts, websites etc.) We become known, some becoming famous/infamous within the SL community. I cannot help feeling that you greatly underestimate the complex social ‘communication’ system that exists in virtual worlds. Many veterans will tell you that their memories of SL and their social interactions are as real if not more real than RL.

Nov 2 2016 at 9:39pm

Thank you Judith and Russ for a great podcast!

It would have been so interesting to dig deeper into the discussion regarding engineered/programmed empathy as a major player in future branding and political communication channels, the risks and opportunities that this poses, not least in terms of possibly ex ante decrease individuals’ expectation of each other’s authentic intention to truly connect and feel for each other (Sherry Turckle writes a lot about this).

Russ, in terms of VR, avatars and signaling – it is long overdue for you to get Jaron Lanier to your podcast; he invented the VR-term! ( you need to get a physical interview with this guy). As a bonus, you could discuss the 2nd Machine Age/Luddite debate and his proposed solution for a sustainable 21th century economy, ending the big data, “off the books” shenanigans and creating a proper 2-way network for all available pieces of information.

Jeff Enes
Nov 3 2016 at 2:04am

Thank you for another great EconTalk episode!

The discussion about signaling and interviewing made me think about the popularity (and some might say success) of online dating. I grew up during a time when dates were found at a bricks and mortar setting, such as the local bar or grocery store. But apps like Bumble, an online dating platform, may be a more effective way to meet people; there is less information overload and perhaps better matching through looking at an online profile first than a chance encounter at the bar.

Kevin F
Nov 3 2016 at 9:09am

Great conversation, but I strongly disagree with Judith’s assessment re the relative dangers of government knowing too much and corporations knowing too much.

If the government knows I have guns, it can take them away. If it knows my religion or race, it can round me up. If it knows I have dissident views, it can assassinate me. If it knows my spending habits, it can expropriate my wealth.

If a corporation knows all of the above about me, all it can do is serve me an ad tailored to my interests.

don l rudolph
Nov 3 2016 at 2:52pm

On the question of who we should fear more with our personal information government or corporations, I would say government. Government, after other methods of coercion have failed, can coerce you with a gun to the head. Corporations must still persuade you.

On the question of signaling and fashion being used as interpretations of personality. Russ saw a person with an out of date tie as a person who was not paying attention. I see someone who follows trends as one of the sheep, someone who needs to be liked. When I see someone out of fashion I think he is saying, ‘Its not my job to please you, it’s your job to please me.’

Judith Donath
Nov 5 2016 at 9:51am

Re saveyourSelf’s question about whether there is a literal arms race in the evolution of communication “I wonder if Judith meant “arms race” in the literal sense. I wonder if she was making a causal argument that advances in communication are potent determinants of natural selection.”:

Yes, this is the premise of costly signaling theory (as developed in biology by Amotz Zahavi in the 1970s and formalized by Alan Grafen in 1990). It can be beneficial to deceive (to falsely signal that one has some quality); such deceptions, though, harm the receiver. If a signal is deceptive too often, the receivers cease to heed it, and a thus a more reliable signal needs to evolve. Reliable signals are affordable/possible for an honest signaler to produce but too expensive/impossible for a deceptive one. Some seemingly extravagant traits, such as the moose’s immense antlers, the gazelle’s stotting behavior, etc. are believed to have evolved because they reliably communicate the animal’s quality.

Some good overviews are in:
Johnstone, Rufus A. 1997. The evolution of animal signals. . In Behavioural ecology: an evolutionary approach, (ed. John R. Krebs and Nicholas B. Davies):155-178. Oxford, UK: Blackwell.

and in:
Maynard Smith, John and David Harper. 2003. Animal Signals. Oxford, UK: Oxford University Press.

When we’re talking about human communication (and some animal communication, too), much of the evolution is cultural – a quite different process: much more rapid, often deliberately directed, based on imitation, not genetics….

Judith Donath
Nov 7 2016 at 6:48am

Re: concern about privacy from govt’ or companies (in response to Kevin F., don I rudolph)

My focus on the risks from companies gathering increasing amounts of data about us is not because I think this poses greater danger than government surveillance could. Indeed, I agree that the government – with its ability to police, to imprison, to accuse and execute — has the potential do far more harm than a corporation does.

Yet, for those of us living with a benign relationship with our government, this harm is only a potential, not an immediate reality.

But the harms from commercial entities knowing an enormous amount about us are immediate, and receive relatively little attention. With the increasing instrumentation of our lives, highly detailed models of our preferences, habits, needs and vulnerabilities have become available to marketers hoping to keep us insatiably needy, anxiously desirous of the next new thing. Although today’s targeted ads are clunky, obvious, and sometimes hilariously misdirected, we are poised to enter an era of highly persuasive, personally targeted marketing.

The risk here isn’t obvious, as an oppressive government persecution is. Instead, it’s the risk of pleasant, companionable influence we face here, the imperceptible molding of our opinions, needs and worries – Brave New World rather than 1984. Note, too, that the line here between commercial and political blurs – for campaigns are marketers too.

A repressive government can use social media and our private information to very bad effect – but they do not need this data to silence and terrorize. The important thing here is to ensure that we don’t have that sort of government.

Andy McGill
Nov 9 2016 at 9:31am

1. It was briefly mentioned, and worth far more discussion, but signalling is extremely prevalent in natural selection and brings us so many extreme species like peacocks and flowers and so many other species. Probably in humans too.

2. I find it so interesting how in almost every public lecture by academics, the academic speaker has to make a joke poking fun at or otherwise denying conservative politics before departing from stringent PC rules, thereby signalling that the speaker is a card-carrying liberal and is trying to make a profound statement, and not an “ignorant conservative”.

Robert Swan
Nov 9 2016 at 5:26pm

Interesting conversation, but I think most of the stuff about “signalling” was silly.

A signal requires a transmitter and a receiver, so why the exclusive focus on the transmitter? Dr Donath seems to see signals everywhere, but how is she so sure they are signals at all? Maybe she just has an oversensitive receiver?

In ancient Rome the haruspex examined the entrails of a bird to read the future. The unfortunate bird didn’t have much say in the matter. Likewise, it appears that regardless of what clothes I wear, Dr Donath would be able to divine a signal from it (I’m sure not wearing any clothes would send a louder signal). No, it’s not signal; it’s noise.

Perhaps I’m an outlier — there’s only one person whose inner thoughts I truly know — but I wouldn’t have the foggiest idea which ties were in or out of fashion. A couple of years ago I did have a fashion-aware moment. I was having a beer with a friend at a city pub and suddenly noticed that nearly all the men were wearing shoes that must have been 5″ too long for their feet (or they were all mutants). Apparently “bozo the clown” shoes were in fashion. I would have laughed, but stifled it in deference to my friend who was up with the fashion himself.

I’m only saying that I do very little “transmitting”, not that I think I can’t be “read”. Sherlock Holmes might get it right; mere mortals mightn’t do so well.

There has been some less subtle signalling going on lately. Yesterday’s US election result showed that the pollsters’ receivers weren’t very good. I have a feeling they might have been closer to the mark if, instead of asking “Will you vote for Mr Trump?”, they had asked “Do you think the British were right to vote for Brexit?”

Jesse C
Nov 18 2016 at 3:24pm

Great podcast, Russ and Judith!

@Judith –

Yet, for those of us living with a benign relationship with our government, this harm is only a potential, not an immediate reality.

Different people feel that their relationship with the government is varying levels of “benign”.

But I see what you mean – Moving forward, I feel slightly better having the government building a full characterization of me than I did a couple of weeks ago. Google, Facebook, Twitter, Youtube? Not so much. I can imagine getting censored, getting my personal information exposed due to an employee whom my political ideology offends, etc.

On the other hand, you use phrases like “horrible, horrible things” and “silence and terrorize” when talking about bad things done by government. You fall short of saying so, but it seems as if the subtler bad things a government might do, armed with an intimate, “big data” view of its citizenry, are less alarming to you.
With private corporations, the bar for what seems to worry you appears to be lower (e.g., a big corporation doesn’t need to collect data about people to commit genocide – but I can’t think of an instance where it’s happened). You appear to hold governments and corporations to different standards.

I could definitely just be misreading your tone.

Brian Mason
Nov 21 2016 at 8:49pm

When discussing privacy Donath is typical as she focuses on privacy from corporations. Her example is it doesn’t matter how much government knows about you if it is intent on killing you.

That is the case when government is completely corrupt. To become completely corrupt government needs information on citizens.

Nov 23 2016 at 11:43am

An excellent podcast.

It reaffirmed my decision to not participate in Facebook, Google+, LinkedIn, Snapchat or any other social media. Also, I only regularly comment on Econtalk and one other forum. Personal interactions remain much more valuable in establishing professional and social relationships.

So Russ, given this and your stated dilemma in the podcast, where are we having lunch the next time that I am in Palo Alto?

Comments are closed.


EconTalk Extra, conversation starters for this podcast episode:

This week's guest:

This week's focus:

Additional ideas and people mentioned in this podcast episode:

A few more readings and background resources:

A few more EconTalk podcast episodes:



Podcast Episode Highlights

Intro. [Recording date: September 13, 2016.]

Russ Roberts: When I recently interviewed Adam D'Angelo of Quora on EconTalk, he mentioned that he had spent a year at the Sociable Media Group, and that it focused him on the importance of signaling. How do you see the role of signaling in online identity? And, is it any different than our face-to-face encounters?

Judith Donath: Yes. I want to just give a little background about what we mean by signaling here. It is a little different than the everyday use of this word. Signaling theory actually comes from biology. And it's--if you think about how we perceive the world, a lot of what we want to know about each other, we can't directly perceive. Are you nice? Are you going to be nice in the future? Do you like me? Instead we rely on signals, which are perceivable indicators of these otherwise unperceivable traits. However, for the person signaling, it can be profitable to deceive: to say that they are nicer, better, faster, smarter than they actually are. And this--the tension between these two, because being deceived is harmful--this tension has brought about, in the world of evolution a arms race in communication that has shaped a lot of the animal world. It's shaped a lot of what it means to be human, our culture and technologies. It's particularly important online. Because, if we think about how we really make sense of the world, it's not just signals, delivered communication, that we use to make sense of other people. But a sort of broader class of things that in this theory we'll refer to as cues. These can be anything. Some things are ways people deliberately present themselves; but others, whether it's their way of walking, how they look, their tone of voice, etc., may be things that weren't meant to communicate, but we still pick up on them. These [?], we see a mix of these two. But it's only the signals that are really open to deliberate manipulation. And somewhat by definition. If you view something to deliberately manipulate how others perceive you, then, by definition it's a signal. You look at how what we can see of others online, almost everything is signal. There's--you can't directly see if somebody is tall. You only can see if they have written, you know, 'I am tall.' Or if there's a photograph, but it's intended to show them that way. And so, understanding the dynamics of signaling is really, really important in understanding how online communities work. A lot of the theory, in looking at signaling, is about trying to understand, given how profitable it can be to lie or to deceive or to exaggerate, what keeps communication honest enough to function? Because if everything was made up and nothing was true, there would be no reason to pay attention to anything. It wouldn't give you any new knowledge. And so, that whole dynamic of how a equilibrium is established that makes communication honest enough to function is the focus of signaling theory. And that's why it's so important to understand the online world, where almost everything is signal. And how we design different spaces changes the economics of how reliable they are.

Russ Roberts: Well, let's talk about some of the ways that we signal, and face-to-face. So, I've been thinking about, recently I bought a watch. A very 19th, 20th century thing, I think, I guess, to do. And I was aware of the fact--I had to think about it a little bit--that for a man, a watch is one of the few ways that we can signal in a face-to-face conversation something about ourselves. There's clothing. There's facial expression. But a watch has some information. And of course, it may not be honest information. I might decide to splurge and buy a Rolex; I'm not interested particularly interested in that signal but some people are. And of course they do that with other ways as well: they buy a certain kind of car, live in a certain kind of house. So I'm always thinking, as an economist, about, to some extent about the way our verbal and visual cues help you identify more things about me. And of course I can lie. Now, what do we have available online that has the equivalent of that? Because it's not obvious at first that there's much that I can signal online. So, what are some of the ways that I can signal online?

Judith Donath: Well, first of all, even in the simplest of interactions, there's the words that you use. Simply what sentence you type. What kind of grammar you use. What kind of vocabulary you use. What did you say? What kind of site have you decided to show up on? There's a big difference in what I'm going to guess about you if I've encountered you on a forum for new mothers or in Forchance[?], which is a forum for misbehavior and trolls. There's--especially in sort of today's online ecology, there's all kind of ways of signaling. If we look at something like Facebook, how many Friends you have is a signal. Who your Friends are. How you respond to them, how you interact with them; whether you use trendy little emoticons or whether you write in full paragraphs tells me something about you; the kind of pictures you put up online tells me something about you. So, there's an enormous number of ways that we both try to influence how others see us and how they give others the fodder to try and interpret what they believe about us.


Russ Roberts: Let's talk about one of the frustrations that I find about online conversation--on Twitter, in particular, which I often enjoy and actually can't believe I'm spending any time on it. You know, someone will accuse me of signaling. They will say, 'Oh, you say that because...'. And I'm thinking, 'Are you out of your mind?' And in real life, the way I could say, 'Are you out of your mind,' the reaction my face makes would tell that other person that they had made a mistake. Say. Of course, that person who wrote that maybe isn't so interested in finding out what I really say. They are just there to poke and make trouble and just express themselves in a particular way. And I find myself struggling to be authentic in that arena.

Judith Donath: Right. So, one of the things--there's a lot of interesting things in what you defend. One piece is, online, a lot of our actions can be for a variety of different audiences. And one of the things that makes online communication very tricky is those audiences are invisible. On Twitter in particular, they are not only invisible but very confusing. So, if you and I and 6 other people are at a dinner together, and you say something, you may be directing it at me but you are also aware of the other people who are listening. And that shapes how you act and what sort of things you might say or want everyone to know, versus just me. But we are all aware of who is there. In some online conversations you may be aware of who is there even if you are not seeing them. Something like Twitter, it's incredibly confusing because we all have different sets of followers and people who we are following. So, any kind of conversation is actually between not only people whose audiences are effectively invisible, but everybody's audience is completely different than anyone else's. So it makes for a very peculiar way of performing.

Russ Roberts: Because--I guess it's kind of like walking into a dinner party, and then, for the second course you are in a different room with a different group of people. Maybe some overlap. But the topic hasn't changed and you are still having a conversation. And it's just a little bit weird, I guess, for those of us who grew up in a different place.

Judith Donath: Well, it's even weirder than that, because it's like being at a dinner party where you see a certain set of people, but some of them are deaf[?] and the people who are actually listening to the conversation are listening in through invisible windows. Because the people you follow aren't the same as the ones following you. So, that's a particularly strange kind of social dynamic that goes on in Twitter. But the other thing you said that's quite interesting is, you said, someone will accuse you of signaling in some way. And that's one of the interesting dynamics about signaling, both face to face and online. It's just generally interesting. Because, as I said in the beginning, we read all kinds of information from people, both--under the umbrella as cues, some of which are signals, which they deliberately or in some cases in an evolved sense, are meant for communication. Others are things that we read from them that weren't meant for communication. And so that only signals are manipulative, that's by definition. And so when we don't think something is a signal, we think it is being done, as you said, in an authentic way, we tend to conflate the notion of authentic with meaning 'Done for reasons other than signaling,' then it's believable. Because it wasn't done to manipulate. So, a lot of how we act, especially when we are signaling something, is in some way to make it appear as if we had other reasons for doing it--

Russ Roberts: Sure--

Judith Donath: So that it will seem more believable. Because authenticity--we think of authentic as the things that people do for their own sake. And so, that's part of the whole, um, manipulation process: is, how can you make things seem like they aren't signals? And a lot of the interpretation is about: Is this person doing that or saying that or wearing that because that's how they really are? Or are they trying to make an impression?

Russ Roberts: And of course we're always trying to make an impression in some abstract, subconscious sense. But what I find amazing on Twitter--and I guess, sometimes even in real life, is someone will accuse me of something--being the pawn of some special interest, hating some particular group. And I want to just yell, 'You don't know me!' And of course, I know that already. Why do I need--why do I have an urge, why would I think for a minute that, despite all that I have written online, written in print, said over the year in this program--why should I be surprised that people don't know me? But it hurts. There's something--and then I watch other people, when they are insulted, say, on Twitter--and they respond with this virulent vulgarity, say, in response to an insult. And I always wonder, drop them a quiet note and say, 'Chill out. Don't feed the troll.' But then I notice my urge to do the same. I don't do it, most of the time. Maybe ever. But the urge is so strong. And it just--I'm not sure it's a very healthy place. I don't know.

Judith Donath: Well, I don't--part of what happens is that online a lot of the cues--well, a couple of things. One is that a lot of cues that we have about the humanness of others are missing. And so, a lot of times people are responding to something someone said. And the other people, and the audience, is very, very disembodied. Face to face, we don't tend to just unleash whatever angry responses we had. For a couple of reasons. On the better side of our nature is we see a person, we think, 'I don't actually want to hurt you. I don't want to cause you harm.' Or, I can slightly say, like, 'Hey, stop it. I see you. Back off.' And we're done. On the other hand, even if we really want to, a lot of times fear will stop us. The realization that if you just tell everyone exactly what you think of them, and it's fairly negative, somebody is going to smack you down, really badly. Online there's no fear of physical retaliation. So, people can feel free to unleash their inner angry self. And there's also no sort of social sense of recognition of the humanness of others. So a lot of the bad behavior is because of that.


Russ Roberts: Some of it comes from anonymity. And you point out that there's big gains from allowing anonymity, and also some losses. And for me, one of the losses is that I think even when there's no fear of physical response when I know who you are, I think it makes it less likely that you will be discourteous or cruel.

Judith Donath: Yes. But that phrase, 'Know who you are,' is a really interesting one. Because--and that's part of the--it's a big part of the current debates over real names: Should everything be done under your real name? Should we allow anonymity? Or should we allow pseudonymity? And I think the--there's a tendency to kind of fall back on real names and real identity as sort of the easy way of asking people to be responsible. And it can be appropriate in some settings. But certainly the fact of knowing somebody's name doesn't actually tell you all that much about them.

Russ Roberts: True.

Judith Donath: Whereas, at some level, what you really--and it has other privacy implications. There's certainly a lot of discussions. One of the things that is nice about a more anonymous online role[?] is people can participate in discussions that they don't want everyone to knows them to know they were in or saying. You can have sort of different facets that way. But, on the other hand, you don't want to have conversations with empty ciphers. And so, I think one of the pieces that is still very under-explored online is what is called 'strong pseudonymity,' where you have a lot of information about someone but it is not necessarily tied to who they are in the real world. But you know what they've said, you know their history, you know their opinions, within a particular space. And so you can feel like you've gotten to know them in that context. You may have more information about them, but they retain some privacy.

Russ Roberts: It's funny: names are somewhat overrated, I suppose. The fact that I don't--that I think, mistakenly, that Middlemarch was written by someone named George Eliot when in fact that was not the author's real name, and in fact the author was a woman, I guess is kind of irrelevant. I know a lot about that author, even though I don't know her real name. Or at least I think I do through her book. So, in a way, you are right: it's something of an illusion. But there's something of an accountability there. Knowing that people know your name--maybe it's the fear that there will be some physical retaliation made possible once the name is known. But I also feel it's a psychological feeling that you've identified yourself, and stood out in some way.

Judith Donath: Right. The difficulty is that there are a lot of other forms of retaliation that could come to people through using their real name. And some of the fear is, you know, if you have a discussion group about, you know, some health concerns, is the retaliation going to come through losing your insurance?

Russ Roberts: Right. Sure.

Judith Donath: If you have views that are unpopular, are you going to lose your job?

Russ Roberts: Or your friends?

Judith Donath: I think you should be able to have--yeah--you should be able to have these discussions in certain places and not have to have every single forum in which you participate rely on tying itself to who you are in real life and keeping you accountable there. Because then you end up with people sort of being boxed into just saying very, very bland things about unimportant topics.


Russ Roberts: Let's shift gears and talk about a more positive aspect of online identity, which is the feeling of, at times, being part of a community. So, I think a lot about my listeners, sitting out there right now listening to this. And there are very limited ways to feel they are part of something as a regular listener to the program. So, if I do a live event, people will come just because they want to see, face to face. They want to see what I look like, they want to see what the guest looks like. They want to thank me--or yell at me--whatever the case may be. But it's amazing to me how much time we spend online. And how much we crave brick-and-mortar, real flesh interaction. And I don't think we've done a very good job creating ways for us to interact online. And I'm curious what your thoughts are, in your study of design. Do you think there are some sites or opportunities that are waiting out there to improve that?

Judith Donath: I think some of the impression is because a lot of the places where people have come to get to know each other very, very well and very intimately are private spaces. But when we sort of go and observe, what are people doing online, we only see the public spaces. And those--you know, it's the same as being, you know, sort of out on a city street. You don't have incredibly close connections to the people sharing those streets with you. So, on the one hand, I think part of what are you are saying about people wanting to be face to face does come, in that sense, that both--there's an energy that comes with being people in the same room. Part of it I think is that sort of cue element--that there is just so much we read from other people, face to face, that we just don't get online, that just being, talking to someone and sort of seeing eye contact, seeing the minute reactions, seeing how people react to each other. There's so much rich social information in those experiences that relatively little online gets close to that. So the places that seem--people say, 'You know, I actually do experience this online,' are in game communities, where people have been in gaming guilds together and been part of a group, and they do raids[?] with others; and there's a lot of that activity and behavior and they get to see how someone else acts under dire circumstances. And they say, you know, like, 'Those people, when I met them face to face I felt like I already knew them.' So it's not necessarily not just about the design of the interface, but what are the people, how are they participating together? I also think that, in terms of design, there's a tremendous amount that exists about other people and how they've behaved and how they've interacted that we make very, very little use of in our design of online spaces. And I think through visualizing some of that--making that information a lot more visible--I think we could create richer spaces that would give us kind of that intuitive sense about who the other participants are that we get when we walk into a room, but we don't when we just start to look at a forum.

Russ Roberts: Of course, one of the challenges is the--sitting at that dinner table with 10 people is not the same as sitting in a stadium with 50,000 people. You can't be intimate with 50,000 people. You can't feel those social cues and nuances and facial expressions that you are talking about. And inherently I think we just have a craving to experience that intimacy and shared experience. Just for the same reason that we go to a movie theater rather than just watching at home. There's just something different about it. It has a different texture. But it has a much different texture when it's conversation around a table rather than back and forth in a comment section.

Judith Donath: Yes. But I would argue that I think it's going to be possible to make some extraordinary interfaces. For instance, years ago in the Sociable Media Group we started doing some work with visualizations of Twitter. And if you go to that stadium and you look at all the people in the crowd, and you can see what they're wearing and their facial expressions--I love sitting around and people-watching in a cafe, you know any kind of crowd. Because you can sort of start to feel like you know something; you can make guesses about people; you can imagine you understand some backstory to them. When you think about what you're actually seeing, you are seeing their face, you are seeing their clothes. How does that compare to something like Twitter, where if you sit down to take a look at somebody's Twitter profile, you could see 4 years' of their commentary on all kind of topics--what they've had to say about all kind of things, about politics, about sports, about their breakfast. There's a really rich archive there. But, right now it's not in the kind of form that you can take a look and get a quick glance, a picture of that person. And so I think some of the data visualization techniques we need to think about is: What is a way of taking that information and making it into a very intuitive portrait almost, in terms of both the frequency of how they interact, what are the topics that they talk about? Maybe some of that has--and there's all kind of ways of approaching it. Maybe some of it's how much do they talk about themselves and their breakfast, versus how much are they a sort of policy-oriented analyst? Or, do they talk about the weather? Is it slang? There's all kind of ways of coloring it. But I think the potential to being able to look at say a thousand profiles on Twitter as a very richly detailed crowd shouldn't be impossible.

Russ Roberts: That's a great point. In fact, I never thought about it before, about how sterile most profiles are. They tend to ask things like where I went to school or what your profession is, how long you've been there. But the opportunity, say, to generate 10 random tweets, or my 10 favorite tweets, or whatever it is, would be a much more interesting--

Judith Donath: Right, or even just sort of analysis of the trends in your tweets. You can have something that would be really rich and nuanced about individuals that way. And something like Twitter is a nice example. It's a public forum; people can be anonymous, or they can be named as they choose, so it's not like you are revealing something secret. But it could just be an extraordinarily rich kind of crowd-watching scene. Which it isn't, quite yet.

Russ Roberts: That's true. In fact, for me, the cafe and the faces going by--it's not so much guessing who they are, what they did, where they are headed or whether they are happy or not. It's just a feeling you are part of the human experience in its diversity and richness and bittersweetness. There's something about a city that isn't captured by sitting alone in your room looking at your Twitter feed.

Judith Donath: Yeah. And I think again, coming back to signaling, I think that it is--one of the things we just don't really recognize about how we spend our attention. That is, both an enormous amount of what we do is signalling: some kind of making a claim about who we are, our opinions and attitudes and status and affiliations are. And we are kind of doing that any time we are out among other people. But it's also that I think an enormous amount of our mental processes really are attuned to reading those signals from others. So it's not just that we are constantly trying to figure out, 'Is this person lying to me or telling me the truth?' but that we are in a constant state of evaluation of the people around us. Of, sort of, who they are; how do they fit into our understanding of the shape of culture--where do they fit, what kind of behavior do we expect from them is something we are just naturally very, very attuned to. So, when we're in a situation like a cafe, that whole part of our minds just kind of richly [?] away.


Russ Roberts: Yeah. Adam Smith said that man actually desires not only to be loved but to be lovely. And we talk a lot, that sentence a lot, on this program; and our desire for respect from others, and of attention, and feeling important. And in many ways, Twitter and Facebook and others appeal to the worst side of us. They encourage us to constantly counter[?], 'Can anybody comment on my post? Did I get some more followers, some more friends, some more likes?' And I think it's really--there's an addictive aspect to that, that most of us are very aware of. But it's very hard to fight. And sometimes it's probably better to just "be yourself," whatever that means. But what I mean by it right now is just, 'Step away from that urge to be noticed, to make an impression, to be thought of as something, whatever it is.' Just not worry about it for a while.

Judith Donath: Well--but it's--I'm not sure that that's actually such a bad side. Some of it, it can be bad. But I spent many years living in New York City, which I love. One of the things I see people find exciting about cities like New York or Paris is that people often take the effort to dress to go out: That the street is really seen as a performance space.

Russ Roberts: Yeah. It's true. Those of us--

Judith Donath: And, the same part of the mind that says, 'I've posted something on Facebook; it's not what I'm wearing,' it's: Is this piece of information that I said, do people find it interesting or provocative or something? How did they react? So, instead of choosing what shoes you are wearing, you are choosing some kind of statement. It might be personal, it might be political. But then you care about how people react. That's how you are, in sort of the City of Facebook. You know, it can be distracting. But, that's how we figure out that balance and how we figure out the right interfaces. But I think it's a very--it comes from a very similar part of our psyche.

Russ Roberts: That's a great point. I'm just--I've been thinking lately about how judgmental I am and whether that's always a good trait. And I think many times it is; and many times it's not. But my urge to judge, constantly as a human being, I'm sure it's built in--it's hardwired. And I have to say, as a non-New-Yorker, any time I go to New York, I'm very aware that I'm not playing by the rules in dress and swagger as the other people there. So it's an interesting--it's a big stage, for sure.

Judith Donath: Right. And part of it also, I think, is that one of the things people also forget is that there are innumerable sort of different sets of affiliation and status that people work within. And some are places where they are really comfortable or they really care; and they may be--you could talk to them; they say, 'Yes, this is a place where I care about how other people think.' But we often do end up in places where we are uncomfortable: we feel this isn't where we perform our best. Or, you know, we don't agree with the mores of the place and we still feel judged. So, what can work very well in some circumstances becomes very painful in others.

Russ Roberts: That's a great point.


Russ Roberts: I just want to add two more things about face-to-face versus not face-to-face, get your reaction. Then we'll go on to a different topic. Because I think about both of these. People--you know, we are doing this interview over the phone. I'm using Google Voice; you are, I think, a landline.

Judith Donath: A landline, yeah.

Russ Roberts: And, in the summer, when I am out in California, I do a lot of interviews face to face. With a slightly different set of equipment. But it's very different to do face-to-face versus over the phone. And everyone assumes that face to face is better. Because it's more intimate; you can see the cues that we've been talking about, the facial expressions. You can read the body language. But in my experience, it's not always true. And it's often the case that over the phone is better. And I've been tempted at times to say, 'I know I'm 400 yards away from your office, but I'll just call you on the phone,' instead of coming face to face. Because we're strangers. You and I have never met. If I had said, 'Oh, I'm in town; let's do this in a studio somewhere,' there's a whole different set of challenges. That's the first thing I wanted to make that observation, that as big a fan as I am of human contact, I think it's not always the most productive thing. And though I said earlier that I love taping EconTalk live, and people like coming to it, I think I'm a different kind of interviewer when I'm in front of a live audience. And there are many good things about that; and there's not so good things about it. So, I just think it's fascinating how, it's not just, 'Oh, you can see the cues.' It changes everything.

Judith Donath: Mmmhmm. Well, part of it is being face to face has so much info that it can make thinking difficult. If you look at studies of eye contact between speakers and listeners--people who are listening, they tend to look at the person they are listening to. It's not that--listening isn't cognitively as hard, and they are scanning the other person's face for all kinds of cues about, you know, are you saying this sarcastically or straight-forward or how do you feel or are you leaving something out? But when we are speaking, we tend to, unless we are trying really hard not to, most people tend to look away from the person they are talking to, to a large extent. Because looking at another human face and trying to form a coherent sentence both are too cognitively difficult to do at the same time. And so, when you are interviewing, there is so much--that even though we think of it as natural, there's so much that goes into being in the presence of another person. And particularly a stranger, when you are trying to think, that it can make it a lot harder.

Russ Roberts: That's a great insight. I really never thought about that part. The part that I'm always aware of is that while you are talking, you might be imagining that I'm listening very closely. And my listeners probably think I am. Sometimes I am. But a lot of times I'm looking down at the clock to make sure that I'm going to get in the stuff I want to get in within a certain time. My mind's racing, because I'm trying to figure out: Should I respond to that? Or should I just let her keep going? And other times I'm looking at my notes, thinking, 'Should I move to this topic--I wonder what topic I should go to next?' And if we were face to face, you would go, 'He's not listening to me.' So, I sometimes tell my guests, when I'm face to face, 'If I'm not paying attention to you, don't mind me.' But of course that's a hard thing, a hard bit of advice to follow.

Judith Donath: Right. Or even when you're at a restaurant with someone: you need to see what time it is, but they are telling you, like you have to be somewhere at a certain time, but they are telling you some, you know, very personal story; you think, 'I can't look at my watch right now.' You get so distracted thinking, 'How am I going to look at my watch? I don't want to insult them. I just need to do this.' So, yes, there's just a lot of rules, face-to-face conversation there's so much to pick up on there. And that was--part of the--great belief in the early days of online conversation was that, well, because it was text we wouldn't have to worry about any of this; and we would just see, you know, truly, you know, deep philosophical thoughts from everyone. But that turned out not to quite be the case.

Russ Roberts: And of course it takes longer to get a glimpse of your phone. So, one of the advantages of having a watch is that if the person who is telling you that intimate personal story of heartache, say, looks away to control their emotions, you can get a quick look at your watch, that you can't get at your phone. Crazy, crazy thing.


Russ Roberts: What about virtual reality? What do you think is coming there? And, you write a little bit about massive online games, where people bring an avatar to an experience that has very limited emotional content right now. Where do you think that's heading, in particular in Virtual Reality [VR] generally?

Judith Donath: Well, a lot of what I had written about with avatars--and this can play out even more so in virtual reality--is, a lot of what is important is how well matched are the inputs to the behavior of the avatar to what the person is doing. So, VR, where you are losing, and it may be sensing where you are, it can sense your face. And that may be jiving[?] an avatar, is socially more promising than, you know, some of the earlier systems where you could have a complicated-looking avatar, but you were giving it very few of the directions how to look. So, if people think back on some of the early days of, you know, graphical social media, where, you know, an avatar would be smiling but that had nothing to do with any emotion you were feeling. So, my interest with a lot of graphical interfaces is, if you have any visual representation of a person it's going to express--we're going to read all kinds of things into it. Are those things being put there because they somehow reflect the person it stands for? And they are deliberate? Or, are they artifacts of the design, and misleading?

Russ Roberts: It reminds me--I'm not a big emoticon guy. But every once in a while, I think, 'Oh, I'm going to throw one in,' say, in a text to my children. And it's something that maybe they've done or said that made me happy. So, I am going to pick a smiling one. But a smile is not a smile. A kiss may be a kiss, but a smile is not a smile. There are so many subtleties, different kinds of smiles. And I go through the 12 or so, or 8, whatever it is on my phone; and then none of them match the kind of smile that I want. So it is a very--it's a subtle thing.

Judith Donath: So, that's one of my pet peeves, graphical emoticons. Because I think that the original emoticons that were made with like a colon and a parenthesis, I think they served a really useful purpose in online communication--is that they basically extended the set of things, like, exclamation marks and question marks. No one doubts that you can put an exclamation mark at the end of a sentence. It changes how it reads. A question mark changes how it reads. And some of the simple emoticons I think were very useful for doing that. It could help you understand: This is sarcasm. Or, it's meant warmly. Or, I'm trying to express sorrow here. When, in the online communication space, when it was really so, this hybrid between the informality of oral communication and the formality of written things, it was really useful to be, to just have more punctuation. Once you start moving into the graphical world, where they are actual faces, then they start being these little beings that you are sticking on the end of the sentence. And the appearance of them, is, I think, distracting and distorting. So, I think that's definitely a step backwards.

Russ Roberts: Yeah. I didn't think about it until now. But someone must have written a dissertation on how punctuation marks have changed over time. Because an exclamation point in 1880 is not an exclamation point in 2016. It's--today it can often mean, 'I'm kidding.' Or irony. And it can mean enthusiasm. Or excitement. But not always. Weird.

Judith Donath: Mmmhmm.

Russ Roberts: Do you think virtual reality is a big deal? I had Kevin Kelly on here; we were talking about the phenomenon. And he thinks it's revolutionary. And I didn't think enough about it; I thought a bit more as a place where people who like games would play. But it seems like something very different is coming. What are your thoughts on that?

Judith Donath: I think it has a tremendous amount of potential because it can be such a rich environment. Being immersed in a space is a very, very strong experience. I think it's going to be quite a while until we really understand the difference between doing VR versus it just being a sort of great big movie. And I think in particular a lot of the interesting pieces will come from the game world, because it's about a much more abstract experience. I also think--again, personally, I'm more interested immediately in the implications of augmented reality. So, in some ways, Pokemon Go, and things like that are going to have to say, and Google glasses didn't succeed, but I think it was a little bit before its time. And when we think about having information superimposed on the world, at all times--I think that's going to be very, very revolutionary. And in particular if we think about having information superimposed on other people. If you think about the combination of augmented reality and face recognition, where, if you could--walking down the street, and any time you see someone--not only--right now, it's like, you have this choice: there's the online world; do we have some sense of information? And face to face where you might--you have rich information at some time but nothing very specific? The ability to match a person to the vast data that's available about them online, I think, is going to have an enormous impact on how our society functions.


Russ Roberts: So, if I'm walking down the street--I'm in New York, say, and I'm dressed really well--which is obviously an example. So, I'm walking down the street, and you come from the other direction, and I see over your head, through my augmented reality set, the last 10 pictures you've shared; your last 5 purchases, your last 5 tweets, who your most popular friends are on Facebook, etc. Or, walking down the street is not as good an example as sitting across the cafe--

Judith Donath: Right. Or sitting in a restaurant.

Russ Roberts: So, instead of just saying, 'Wow, that's a great suit that that person is wearing; it must have cost a lot of money,' or 'That's a really bad color for that person,' or whatever it is, instead I'm going to get this unbelievably rich--in theory--set of information. And it raises an issue that you talk a lot about in your book, The Social Machine, which is the role of privacy. So, right now, I might be able to find out--when I google someone, I find out lots of things about them. Whether they want them shared or not, I'm going to find their wedding announcement; I might find that one of their parents has passed away; I might find what sport they played in high school or in college. So, you get that information, not real time. But it raises the question how much you'd share, and back to our signaling discussion, how much you'd want to share in that setting in the cafe, versus something more intimate.

Judith Donath: Right. So, for now, it's public information that's available about people. But even without getting into, you know, whether you can see my purchases or things we think of as private information, our whole notion of what is a stranger will really kind of disappear. But it also means that the distinction between those who have extensive online presence and very little presence will be manifest, you know, as soon as you see somebody. Plus any of the negative information that's available about someone. It's not necessarily the portrait they would choose, that you see. You see the portrait that is made by whatever service you subscribe to that creates this kind of data about the world around you and presents people from that perspective. So it has tremendous privacy implications. But I also think that ultimately, desirable or not, it's not something people will want to go back from. And so the way we now look back and think, 'Oh, my God, how did people, you know, live without running water? Or bathrooms? Or electricity?' we will look back and think, 'Oh, my God, how did people live surrounded by people about whom they knew nothing?' It will seem very strange.

Russ Roberts: Yeah. They'll read Jane Austen novels to try to get a vague idea of what that must have like.


Russ Roberts: When we think about signaling, one example I think about often is a necktie. Which I think of as a signal as to whether I'm paying attention. So, I know people--and I don't wear a tie very often, so I'm not a very good example--but I know people whose ties are grossly out of fashion. And they tend to be older, people who just haven't changed their tie collection since they were 30 or 40 years old, and now they are 60 or 70. And their tie says: I don't pay attention. Or: I don't care to pay attention. Right? And I think what's interesting about this potential revolution is that the levels and ways in which you can signal that you pay attention--which is something people do care about--whether you are hip, whether you are on top of things, whatever you want to call it--is going to have an incredibly expanded array of ways of doing that.

Judith Donath: Yeah. And I think this notion of hip and fashion, which we tend to think of as very frivolous, is actually a lot more important in our culture than we recognize. Because we really do live in a world in which access to information is one of the key markers of status. I mean, wealth is certainly a very important one; but the whole notion of how do you get access to information, how good are you at assessing information, is a huge amount of how we display who we are. And that's really what fashion ultimately is: It's a signal of how good, how much access to information do you have and how good are you from parsing the good information from the bad information. And it's not just in the world of clothing. I mean, that's the very obvious one, where we think of fashion. But there's fashions in management styles. There's fashions in academic topics. There's fashions in all kinds of things. Slang is a fashion. One way of thinking about Twitter is that it's using news almost as fashion.

Russ Roberts: It's true.

Judith Donath: Where, you know--can you get the most current story out? And are you willing to take the risk of tweeting something in its very early stages, when it might be wrong? So, you know, that's where the cost of signaling comes from. And so, I think you're right that all of this sort of ability to display how up-to-date you are is going to continue to accelerate.

Russ Roberts: But as you point out, if you are reading a book on your Kindle, you can't signal to people how intelligent or hip or arcane your interests are, the way you could with a physical book. It's just interesting how technology has changed that, at least that opportunity, in the other direction. But your point about fashion, it's really I think a deep one. Because it's really tangled up with our identities. Right? So, it's not just--certainly, many of us--again, I'm not a particularly good example of this--but many people, the clothes they wear are an important part of their identity. There are those of us for whom not caring a lot is an important part of our identity. But on so many other dimensions, the things we wear--and now I'm going to use that term very broadly, beyond clothing--the things we wear, meaning the food we eat, the people we hang out with, the topics we follow, the interests we have, the things we read about, the things we have opinions about, the hobbies that we indulge in--those are our identity. It's everything, in many ways. Not just all these things that are on the side, just aspects of me. They are me.

Judith Donath: Right. And all of them are, to some extent, subject to fashion, in how you express yourself, and which particular foods you like have fashions. It's not just restaurant fashion, like there's fashion where kale came into style and is on its way out of style.

Russ Roberts: Oh, no! Good riddance, I say. I think it's overrated.


Russ Roberts: I want to ask you about my 86-year-old father who likes to ask me if he should get involved in LinkedIn, or NoseBook[?]? He's not up to date. He does have Internet access. He does not have a smartphone--he doesn't have a phone. That's a different kind of statement he's making: he finds it "unnecessary." And as his 61-year-old son, I find it frustrating at times when he struggles to interact with the technology that he has chosen to deal with--say, Facebook or a webpage or whatever it is. And I look at him--he's 86, and I'm 61, and I just sort of assume that I'm just going to be like him when I'm 86. Hard to accept. But that technology will have changed so much that it will have passed me by. And then I wonder: My children, who can't understand how semi-inept I am: will it pass them by, at 61 even? 50? 40? Is it that the world is moving so quickly or is it that some people are grown up pre-Internet and post-Internet?

Judith Donath: I think it's that the world is moving very quickly. Because this is--part of what I'm writing about these days is a lot about signaling and a lot of it is about fashion and signaling. But some elements of fashion really function as these kind of social gatekeepers. And consciously or not, I think a lot of social technology--so I have a 16-year-old daughter who is very much into fashion of all kinds. And as I watch her and her friends cycle through technologies, it's both, there's a sense of--and she will not even look at Facebook, because it is too embarrassingly the space of old people.

Russ Roberts: Yeah.

Judith Donath: You know. And she has a 19-year-old brother for whom Facebook is just fine. But that's just--three years, or a generation earlier. But a lot of the interface has also become increasingly opaque, or require knowledge of a previous interface, so that you just have to understand that, 'Yes, that quadrant of your phone has a certain meaning.' There's no particular marker of it; it's just the knowledge that you have. And that's how things like fashion really display, make you display whether you have a particular kind of knowledge or not as a marker of whether you are in a particular group or out, or how, sort of authentically in it you are. And so I think a lot of these social technologies really do have that movement of requiring that you are willing to continuously adapt to the new is what you are displaying by using them.

Russ Roberts: It's interesting that that's a virtue, right?

Judith Donath: A feature. It depends how you look at it. From the perspective of structuring society, it's a feature. From the perspective of inclusiveness, it's a bug.

Russ Roberts: Because it's a burden.

Judith Donath: Right. And it's similar with language.

Russ Roberts: Good example.

Russ Roberts: If you look beyond the interface at how people use these different technologies they have on their own developed a lot of mores and languages that are also about: Do you not have the knowledge to perform this and use this correctly? And if not, that's a marker of your outsiderness.

Russ Roberts: Yeah, it's just not obvious that inclusiveness of that kind is really a--that the urge to be on the inside, whether it's with the latest clothing fashion, technology, slang, etc.--I'm not sure how socially productive that is. Was that your point, perhaps? Maybe I missed it.

Judith Donath: No, I don't think that's my point. I think the urge to divide the world into in-group and out-group and the urge to be in the in-group is pretty fundamentally part of what it means to be human. That's something you can sort of look into neuroscience research and how our brain reacts to those we perceive as in or out. There's a lot of plasticity about how we define in-group and out-group. It's quite plastic. But, we do that constantly. And so an enormous amount of signaling is about claims of affiliation, of being part of a particular group, whether it's a particular class or social group or the set of people who are interested in x or are Trump supporters or hate Trump or whatever. So I guess what my point with fashion is, is that since the development of fashion about 500 years ago, it increasingly isn't important whether you like it or not--that's a different thing--but it's an important marker of in-group and out-group membership in a world where access to knowledge is part of what marks those distinctions.

Russ Roberts: Yeah, I guess that's the positive way to think about it. I like [?] that access to knowledge thing. On the other hand, I feel like you can waste a lot of time trying to figure out what you are supposed to be thinking or wearing or saying that is truly part of human flourishing. But I think you are right: You can't put that genie back in the bottle. You just have to make your own choices. But I get my dad, and I'll be him, God willing, in a few decades.

Judith Donath: But the interesting thing, getting back to your tie, this is one of the things I find very interesting about fashion, and I certainly see this a lot with my 16-year-old, is this notion called 'countersignaling,' where if you have this tie and right now it's very out of style--in my case, my old jeans--my 16-year-old will wear them. And she's walking around in a pair of jeans from the 1980s. And her shoe goals for this year are a pair of sandals that Rihanna is promoting that look exactly like old lady bedroom slippers. And it's either that or Birkenstocks, all of which on somebody who was not in many other ways signaling, 'I am the height of fashion,' would mark you as extremely out of fashion. And it's a very interesting way to signal that you are so fashionable you do not have to worry about being mistaken for unfashionable. And it cannot be copied. That's one of the things, since the earliest theories of fashion, are that, like, the most fashionable, have a new fashion, and then the next below copy them; and make the top ones have to change. And countersignaling is the uncopyable fashion, because if you are not at the height of fashion, you can't copy it. Because people just look at--

Russ Roberts: Yeah, I can't send that signal that way. I have to find a different way to send it. That's very cool.


Russ Roberts: Let's go back to privacy for a second. You said something that I've started to think of as true, which is: We don't want to go back there or we're not going to go in that direction. It strikes me that part of the problem with privacy in its current state is that it's moving in one direction: we are getting less private. Which gives people a chance to exploit us, manipulate us, steal stuff from us, change our vote, our purchase, whatever it is. And at the same time it's this incredibly rich world of customized things for me that I love. And it's hard to imagine we're going to give up one to get the other. And it feels to me like we're just going to move to a less private world. And, in fact, again thinking about my parents, people of that generation, when they see what goes on in our world, that they are not so much part of, it shocks them. And I think in 25 years what we do to protect our privacy may shock us. I don't know. [?] shock them.

Judith Donath: Yeah. Well, part of what makes a lot of the privacy discussions complicated is we also need to articulate who is invading our privacy. Because there is privacy from other individuals--what do we care about our peers or other individual people know about us. There's privacy from corporations. There's privacy from corporations. There's privacy from the government. So a lot of it really hinges on who is using it and how do we feel that the information about us might be exploited or harm us, versus what are the tradeoffs that we get. At some level, I think we're also coming from a time of unprecedented privacy. Because if you look in history when people lived in much smaller communities, you didn't have a huge government looking in on you, but your neighbors knew a tremendous amount about you. And so, sort of the 19th and 20th century urban life was one where so many things had moved to the marketplace you didn't need sort of a constant web of social connections to get your food, or having your house exist. You could just go to one place, make some money, trade that money for other things; and live this very, very private life that really is, I think, from that standpoint, somewhat unique in history. And now we're moving to another unique point where there's a lack of privacy but on a vastly different scale of surveillance, of being watched without necessarily watching others. And so a lot of the privacy concerns aren't so much about our connections with other people but with these much larger groups, about whether it's Amazon or Google or the government, that are watching us and have the ability to manipulate us.

Russ Roberts: Yeah. Power is, and competition, are really central there. There are a lot of things I never want the government to have--because it doesn't matter how nice they are. They are not to be trusted. Powerful people are not to be trusted. It's not in our psyche to be trustworthy with power: that's my view of history and humanity. So I think it's really different, in that--your point about across people versus across organizations.

Judith Donath: Right. Yeah. And for me, my concern is more with corporations than with the government. I think partly because, if a government goes bad, it doesn't really matter what they know about you. Governments have done horrible, horrible things with really little surveillance whatsoever. If they want to strike terror into people, it really doesn't matter if they go after you because they knew something about you or they made it up. They will. And so I'm less concerned because it seems like the issue there is: Is your government for the people or not? And if it's not, privacy or not isn't going to be the crux of making things better or worse, necessarily. But the level of commercial manipulation that we are dealing with is, I think, a much more immediate threat. And one that we haven't really seen before.

Russ Roberts: Yeah, it's so interesting. I like your point about government in general. The first point, I like a lot: that you don't need a lot of information to exploit and abuse people. But I would push back on that and make the observation that the more they know about us, the easier it will be for them to do other things that they hadn't thought of. It's true--the Nazis didn't need a database. They would have liked a database of Jews. But they made one anyway, with primitive techniques; and now it would be a little quicker. It's not that important. The problem is that their ability to exploit--ambitious people who know a lot about us without competition from others is a terrifying thing for me. And that's why I worry about the government. And I think that's why I worry less about corporations, because in general they face competition. I think the issue now, that even people like me who are free market have to cope with, is that, on the surface at least, many of the corporate giants who know an enormous amount about us--the three obvious ones being Google, Facebook, and Amazon--maybe the competitive environment there is not what it was in the past. I think it could be. I don't think Google is going to last forever, or Facebook, or Amazon; I think they can all be destroyed by competitors. But maybe their ability to use their current positions in the face of the actual, the not-so-much competition of right now, is worrisome.

Judith Donath: Mmmhmm. And I think--there's[?] so maybe you weren't worried about. But, you know, I think also, as--getting back to signaling--one of the things that I think is going to change over the next decade or so, is the increasing number of our interactions are going to be with non-human agents. So, whether it's chatbots online, or the descendants of Alexa and Siri in our homes. But lots and lots of things that will, to our minds appear to be appealing creatures that we have an emotional relationship with and respond to in an emotional way. But which are designed to appear as if they have feelings and have a relationship with us, but are really designed for the benefit of the corporations that make them. So I think there's a level of persuasiveness and manipulation that we have not seen before that we will, in the next decade.

Russ Roberts: It's an interesting point. I do think there's a potential for exploitation there. It's done now, without technology. Obvious example for the bank with the slogan, 'We want to be your friend,' or some awful thing like that. Well, they are not your friend. They want to be your banker and make money offering you transactions. But, it sounds good, and there's nice music playing in the background; and we have this emotional response. If there's lots of banks with competition, and there's lots of Alexas and Siris, I like to think that the ones that don't exploit me and that actually do me well are going to be the ones that are going to be more successful. So, to me it comes down to whether there's going to be competition in these spaces. And, we'll see, I guess.

Judith Donath: Mmmhmm. You know, or the same technology, as it becomes more a part of how our government is elected.

Russ Roberts: What do you mean?

Judith Donath: Well, right now, you know, some of the mailings that you get are directed to you; but you don't get a completely different message from every candidate. So, as it becomes a lot easier to--one, be conversant with things that aren't necessarily human but are very good at persuading you, and can talk to you, and you know, say, well this is, you know, this is what I support--I think there's just a level of artificial camaraderie that will be much more influential in a number of realms.

Russ Roberts: I think that's true. Why don't we close with this? I think you and I like to--people "like us" like to think about these things. And we like to think that we are sophisticated. And I won't include you. I'll just talk about myself. Sorry about that. I like to think about myself as "sophisticated." You know, I'm sensitive to these things. I know when I'm being manipulated. I know that Siri is not a person. I don't have any emotional ties to her. But then the question is: But there might be other people, they are going to be susceptible to this? So, are you optimistic that we could encourage lots of us to understand these powerful techniques? Or, are you respectful of everyone's ability to do it? Are you optimistic, or do you think actually that none of us can do it--we are just fooling ourselves and we are all susceptible to it?

Judith Donath: I think we are all susceptible to it. And I think that the attempt not to be susceptible is dangerous in itself. Because, you know, do you want to harden your heart against anything that appears to be vulnerable? You know, it's a little creature, and it's in your house, and it's like really disappointed that you don't actually want to buy the jeans it suggested for you? But, trying to be too cynical, or is to say that you are closed to emotional manipulation--does it close us to be emotional interaction between us and other humans or other animals that actually have real empathy and feeling? So, how to keep that balance is going to be tricky.