Signals, Truth, & Design


Uploaded by Google on 24.07.2007

Transcript:

MEGAN SMITH: Hello, welcome.
I'm Megan Smith from new business development and I
have the pleasure of introducing Professor Judith
Donath, who is here from MIT today.
Judith's work is fabulous.
Every time I get to go to the Media Lab and see something
new that she's done, I have yet another vantage point into
how social media or media communications media work.
Today she's going to talk, as you see, about Signals Truth
and Design, and everybody got the email so I'm not going to
go into detail there.
She runs a socialable media research group, and they work
a lot on the social side of computing graphic design,
urban studies, cognitive science, all these things
pulled together into a lot of interface work that they do
for online communities and virtual identities.
She's known for tremendous social visitation and
interface design.
I see some of you I folks here, so I'm very happy.
You guys are here to see some of the amazing
work that she's done.
And you guys have a copy of her bio so I won't read from
it, but I'm just so happy she was in town and I asked her to
come and talk to us.
And I think there's a lot we can gain at Google always by
looking at how things are really working and taking
another point of view.
And she's done a tremendous job throughout her research in
doing that.
JUDITH DONATH: Thank you, Megan.
So just as a start, not to contradict that, I do a
tremendous amount, as Megan said, of visualization work
interface design.
This talk is going to be kind of light on the design work
though I guess you can all vote and you can say stop this
we'll just talk about design instead.
What I wanted to talk about today was some more underlying
theory about thinking about social interaction online and
how people communicate and some of the ways of thinking
that inform the sort of work that we do.
I'll put in a little bit of some our interface design work
maybe at the end, but this is more of a talk about new ways
of thinking about how people communicate and why particular
designs are interesting.
And what I'd like to start with is introducing this
notion of signaling.
Signaling is a theory that comes from
the world of biology.
The first person to really look at this was a Israeli
biologist named Amotz Zahavi.
And he was interested in some of the natural phenomenon such
as there's Moose that have these tremendous horns, like
these big antlers, and these antlers are very
metabolically costly.
Or you look at some bizarre behaviors in nature.
It turns out that some kinds of gazelle when they see an
enemy, instead of running away as fast as they can which
would seem like the sensible thing to do if you're one of
the fastest creatures on earth, instead what they do is
they jump up and down in place which seems like kind of a
peculiar thing to do.
And he was interested in coming up with an
understanding of why would these phenomena happen.
He came up with this idea of handicap signaling.
Which says that sometimes if what you are doing is trying
to advertise some kind of resource that you have in
great quantity the best way to advertise it is to waste it.
So if the gazelle has a great deal of speed, one of the
things that jumping up and down is
doing is wasting speed.
It's a way of saying yeah I am so fast I can just sit here
and jump up and down use up my energy, use up this time, and
you can go and chase someone else.
And it turns out that predators do not chase the
gazelles that do this because they know that once they start
to take off after them it will be a fast all out energy
losing race that they are likely to lose.
And so, many of these things have turned out
to be honest signals.
That by wasting particular resources is a way of honestly
displaying them.
Now when we look at a lot of ways of self presentation on
the net today, honesty is not really the first thing that
comes to mind.
It's always been a big issue, for instance in dating sites,
where it's just as easy for me to say I am who I am.
I could say I'm the Queen of England.
I could say that I am a 35 year old male bodybuilder.
There's really none of those types of cost they keep things
honest. Now there's various other things as most of what
we see about people's self presentation is not
necessarily a flat out lie.
For instance in dating sites, many people figure that at
some point they're going to actually meet someone, and
some of the truth of what they're saying
needs to come out.
But then there's others who are doing this as a game.
There is the whole bored fraternity brother theory of
dating site.
You know how many people can I get to answer this ad.
So they may be completely imaginary.
So what we are interested in doing in my group now is
looking at signaling theory to get some kind of understanding
of how we can design sites not so that everything has to be
honest but so that, as designers, we start being able
to understand.
The way that if you're a restaurant designer you can
play with different types of lighting.
You can say I want to make a nice family friendly place
where everything is clean and people are fast, and I'm going
to use those bright halogen lights.
Or you could say I want this slow romantic dinner, so
everything's going to be dark.
And yes, if this was a fast food restaurant it would be so
dark you would not be able to see the cockroaches go by.
But considering it's a very expensive fancy restaurant
we're not worried about lighting to see the
dirt under the table.
So in such a way what we're interested in is helping
people understand how they can go about using sort of the
whole range of design strategies for making social
interaction more honest and reliable.
So I'm going to give you a quick background on signaling
theory and then show how it applies to some kinds of
online design.
The first thing is a vocabulary piece.
We can think about any of the things that people use to get
information about others are cues.
Some cues are intentional these are the ones we're
interested in terms of signaling.
And there's what I'm going to call evidence or unintentional
cues which is also a great deal of what we
read from each other.
But the notion is that the signals are things that are
intended for communication and it's this intentionality in
the communication that gives rise to the types of dynamics.

I use a fur coat as an example because a lot of things can be
a mix of intentional and unintentional signals.
So for instance, someone can wear a fur coat to say oh look
at me I am so stylish and not only am I stylish I am rich.
I am wearing a coat that while I could stay warm for $150
from Lands' End end of season sale, I have instead chosen to
spend $15,000 on it.
So it's a fairly reliable signal of wealth.
It's a signal of stylishness depending on the cut.
However, someone looking at them may read these signals
but they also may say that the cue that they see is oh this
person is cruel to animals, heartless and thoughtless.
So that is not an intentional signal but it's still
something that is being picked up on by the
person looking at them.
So here's is an example of the assessment signals I was
telling about at the beginning.
These are costly signals.
For those who are interested, I can give URLs at the end of
more extensive papers about this.
But these are signals because they're some way
of wasting a resource.
This resource can be time.
One of the earliest sort of protosignaling theory writers
was a early sociologist named Thorstein Veblen who you may
have heard of.
He's famous for coining the phrase conspicuous
consumption.
He was a social critic at the turn of the 19th century.
And one of the things he wrote a book called The Theory of
the Leisure Class.
He was interested in if leisure is such a sign of
status, how do you know that someone has a
lot of leisure time?
I don't really want to spend a whole year watching you do
nothing to know that yes you really didn't have to
work for that year.
But so what he talked about was effectively assessment
signals for wasting time.
And his examples were knowing dead languages, raising
pedigree dogs, things like that.
Where there are all kinds of things that take enormous
amounts of time but don't provide any income.
And respectively, these are assessment signals of time.
So it's many kinds of resources you can look at it
in this way.
Obviously, people buy luxury goods as an
assessment signal of wealth.
And again, one of the things to keep in mind is, we'll talk
in a minute about conventional signals which are things where
there is no cost associated with them.
So most things are signals of many many things.
Again, something like a very expensive car may be a signal
of wealth and the ability to waste money in order to show a
high level of income.
But someone buying it might say well I'm really buying
this so that I can seem very very sexy and attractive.
But that quality is not required by being the
owner of the car.
So we have learned by convention to associate it
with that, but it's not necessarily a reliable
indicator of attractiveness even though it may be meant to
be one by association.

Something can be costly in terms of missed opportunities.
So what you can say is something like having multiple
facial piercings is a way of saying, in a reliable sense,
that you are affiliated with a alternative lifestyle.
Because what you have done, is put yourself in a position to
pay the cost of being unemployable in a whole range
or being unpresentable in a whole range of other
situations.
So cost can be in time, they can be in resources, they can
be in opportunities.

Indices, and I will talk in a minute, we'll keep
this tiger in mind.
Indices are another type of signal that's very reliable
because it's impossible to fake.
It requires some kind of natural ability or native
ability that you simply have to have that ability in order
to show the signal.
There's a book called Animal Signals by a British biologist
named John Maynard Smith and he used the tiger as his
example of indices signals because it turns out that it's
a way of signaling size.
It turns out like if you're a really big tiger and you
scratch, you scratch really high up on the tree.
If you're a little tiny tiger, your scratches are
just not that high.
So the height of the scratches on a tree is a reliable
territory marker of height among tigers.
And the reason it's interesting, we'll get into
that in a minute, just keep the tiger in mind.
Obviously some things like having big muscles is a indice
of strength.
It's not necessarily something that weighs strength but
especially if you are showing it off in a particular way, so
there's signals that are amplifiers
of particular indices.
If you're very strong and you walk around in a bikini,
that's a amplification of that indice of strength.

One of the things with these assessment signals they're
costly for the person or animal giving the signal.
They may be costly to assess.
Conventional signals are things that don't have these
associated costs, they are less reliable, but they're
also less costly.
But they're very very common.
Less common in the animal kingdom though they're not
completely absent.
But they're very very common.
A huge amount of human communication, and
particularly our communication online, is in the form of
conventional signals.
And conventional signals are far less reliable.
But what we have, especially among humans, is there's other
ways of keeping them reliable.
In particular, society can add costs such
as punishment costs.
So for instance, I might decide there's a lot of
traffic here.
It would be very very convenient for me to put a
siren on top of my car so that I could
pretend to be a policeman.
A siren is effectively a conventional
signal of being a policeman.
I checked it is really easy to buy them online.
You can just order them.
And I could sit there and say ha I'm a policeman excuse me
out of my way.
But there's such a high punishment cost to doing that.
It turns out that the punishment of impersonating a
policeman is really really not worth it to get home 15
minutes faster.
And so society manages to rebalance
these types of costs.
The understanding that type of dynamics is particularly
important because in the animal world there's very
little ability to fake the more assessment
or indice type signals.
But humans can fake almost anything.
And one of the reasons I like that tiger example is cause
Maynard Smith in his explanation of tigers, it's
not a very humorous book in general, but at the end of his
description, he said now this could not be reliable signal
if tigers could figure out how to stand on boxes.
But they haven't.
But humans actually are very good at doing things like
figuring out how to stand on boxes.
So we always find some way of faking some kind of signal.
There is very very little that you can think of that's a
reliable signal.
That there's not some way, if the benefits of faking it are
high enough, that it's not fake.
So deception occurs when it's beneficial to the signaler.
So the cost of doing the fake signal is beneficial to them.
But deception is very very costly to
the receivers usually.
Some kinds of deception aren't.
We live in a world in which there is almost constant
levels of deception.
I have two little kids, for any of you have children one
of the things you start to realize a lot of what you do
in civilizing children is teaching them to lie.
You do not tell people you think they're ugly.
You do not tell your grandmother what you think of
the present she gave you.
There's just an awful lot of things you don't want to tell
the truth about.
Children are not very deceptive.
We spend a great deal of time teaching them how to lie, when
to lie, what are the right kinds of lies, how to ignore
other people's obvious lies, and things like that.
But in the cases where for the receiver, when they have
deemed it to be in some way costly, there needs to be some
way of changing that.
And also deception is very costly to honest signalers
because what it does is it makes the reliability of their
signal as a whole decrease.
So when the cost to the receivers is too high they'll
cease listening to the signal.
If, for instance, when people first got online most people
assume that when someone said I am a 26 year old buxom woman
who is just looking to spend an evening of love with you,
that this was amazing, that it was great, and they were
really psyched.
It didn't take that long for the receivers of messages such
as that to start to realize that actually it probably
wasn't true.
And that the result there was simply to ignore many many
messages like that.
As we said, another possible thing is if society as a whole
decides that this is costly so increase the cost by adding
societal punishments.
If the cost to the honest signalers is too high, and
you'll see once we get a specific example of these
dynamics, one thing they may have to do is just find a more
reliable signal.
One of the things that we are going to be talking about some
of the samples we'll go into greater depth here, is a
notion of fashion.
And how fashion in many domains is effectively a way
of constantly changing the forms of signals as imitators
come up to constantly try and maintain a
new but honest signal.
The other possibility is that everybody
starts to become deceptive.
Yeah I think one of the things that's really interesting in
our society is, as an example, is all the things that in
previous generations such as having crow's feet the age
marks as cues of age have now turned into signals.
That as the technology has changed and things like
plastic surgery and all kinds of youth enhancing formulas
become much more ubiquitous, many things that had
previously only been cues start to become the signals of
how much time, energy, and pain, are you willing to
dedicate to looking youthful.
So anyway, one of the things that we can do with signaling
is start to explain seemingly irrational behaviors.
I think an example that's very interesting, going back to the
beauty version, is sun tanning.
Sun tanning is a extremely common almost ubiquitous human
behavior yet historically it's had a very
very different meaning.

Do you have a question?
You look like there was some confusion.
It's a beach scene.
But one of the things that's very interesting about sun
tanning is that tanning is a signal of social position very
much related to leisure.
And it's one of these sort of basic stories of how cultural
mores change is at a time when if you were wealthy you were
able to stay out of the sun, when many of the people who
were poor and labored in the fields were out
getting tan and brown.
Having a tan was not considered any
kind of status signal.
As increasing amounts of low status labor started to move
into factories, slowly the cultural perception of the
meaning of tanned skin changed.
And most of the social histories of this say that it
was Coco Chanel who made it very explicit sometime in the
1940s that tanning was now the high status symbol.
Because in order to have a tan, particularly tan in the
winter, meant that you had to have the leisure to go to a
sunny location, and so the whole social
meaning of it changed.
But it's also a very interesting phenomenon.
As time went on, this is a poster from the '60s, it
became something that was very very popular.
People spent a lot of time lying in the sun
trying to get a tan.
Saying OK I have like four more hours on this side and
then I have to turn on the side and spend a few more
hours this way.
Then you'd have this mark that said I've had the leisure to
do this, I have the time, if it's winter especially, I've
had the opportunity to go someplace sunny.
And then between 1980s 1990s some very
interesting things came up.
So first of all, what we see is that we are human and, so
we figured out that way of standing on a box.
It turns out there's boxes to stand on in
order to get a tan.
And the cosmetics industry spent untold amounts of money
both figuring out how to make fake tans that you could
slather on, the development of a huge industry
in sun tanning equipment.
So that you could spend your entire week slaving away in
the office and then go home spend an hour a day at the
tanning booth down the street and come out at least looking,
though it had not actually been any kind of reality, like
you too had this wonderful leisurely life.
But what's also very interesting about it, as I
said, as a rational behavior that's interesting is that
when people talk about tanning when people have been
interviewed what we like about it is that it
makes you look healthy.
It makes you look like you've been outdoors.
And at the same time, it's becoming increasingly
associated with notions of health.
Many medical reports have come out saying this is an
extremely unhealthy thing to do.
This is going to kill you.
If you go tan, you're going to get sun cancer.
Also it's associated a lot with people feeling like they
look young, that they feel healthy, it's going to make
your skin wrinkled.
So one of the questions that you can look at through
signalling this way, and it's one of these things that
there's a lot of academic papers looking at how can we
get people to stop tanning.
There's all kinds of warning ads.
And I think if you look at it as a signal, you can look at
it from a very different perspective that's very
interesting.
Because if you say well actually may be as a signal
there's something about wanting to signal now being
young and youthful and healthy that the idea that this is so
unhealthy is actually beneficial for the signal not
for the person necessarily.
Because it now becomes, in addition to everything else,
it becomes a way of saying I am so confident of my
youthfulness and my imperviousness to melanoma and
all kinds of diseases that happen to other people.
That I can go out and defy all of these odds and lie in the
sand in the sun or in the tanning bed and come out and
look great.
And it's in some ways because there are increased numbers of
warning that it then strengthens this behavior as a
way of showing that you have this enormous excess of the
resource of health.
And so as a way of looking at things, I mean it's a very
very interesting way of trying to analyze this because for
instance if you look at all kinds of prevention programs
whether it's trying to get people from sun tanning to,
I've been working somewhat with someone who's been trying
to do a lot of safer sex work among teenagers, that
realizing that there are often the desire to do something
dangerous because of its riskiness, not in spite of it.
And looking at what the communicative function of
doing that is really really important in trying to
understand how to devise better campaigns for
influencing behavior.
The example I'd like to go into in a little more depth
here because I think it's of particular interest in the
context of Google is this notion of fashion.
And by fashion I don't mean just clothing.
But there's fashion in all kinds of things not only in
our particular blogging sites, the blogs themselves really
function as fashion.
So let's just back up for a second and say
what I mean by fashion?
In terms of signaling the way of thinking about fashion is
that it's something where the form of the signal itself
changes over time though the thing that's being indicated
remains the same.
In particular, fashions are signals of social status and
affiliation, we'll talk that this is in a minute, and of
information access in a very very
information based society.
So you can could look at a lot of blogs and say you know
there's a huge fashion component to blogging where
any one piece of news that's there isn't the key.
It's this notion of saying, on a daily basis now, we have to
link to the right thing.
If I want my blog to be at a certain level of
up-to-dateness where other people will link to it because
it has interesting things, I have to be able to go out and
navigate through this complex world of information and find
the thing that is the right thing to be linking to today.
And obviously, I'm sure you've done a bunch of these here
these measurements and sort of how waves of links move
through the blogging world.
You have a small number of early adopters the people who
find the first ones and there's the ones who go out
and link to something at the very beginning then there's
this big mainstream wave. It is like lots and lots of other
people whose blogs are not read to find the forefront but
just who in order to say yes my blog is actually a display
of my being up to date in the world of information within
like three to five days have the same link.
And then it kind of tails off and then a couple weeks go by,
and nobody is linking to it again.
If we look at it in terms of risk, the risk is much higher
as with any kind of fashion for the
people at the very forefront.
Because whether it's clothing fashion it's saying well did
you just invest in something that is going to be laughable
that is never going to catch on but you were at the
forefront, but you made a mistake about this.
But in the blogging world, the people who are really anxious
to be at the forefront the ones who are putting in four
or five, six hours a day writing a blog have to pick
through all the information they come across and know that
they have to pick something that is
going to become popular.
If you pick something that no one is going to find
interesting, you will start to lose readership.
But it can't have already become popular.
So there's the high cost of finding that information and
the risk in either writing about something that's going
to turn out not to be true.
An example there was a very brief flurry a couple of
months ago, and I was writing about this, of news stories
about how McDonald's interactive games division--
Does anyone know this story?
So it was one these things that kind
of tailed off quickly.
Apparently McDonald's had this interactive games division
that was supposed to be doing simulations within McDonald's
of their future development, where they should be opening
up new branches, et cetera.
And at this game developers conference, on serious games
and simulation, they came in and they said we've seceded
from McDonald's because they are not listening to our
recommendations.
Because it turns out, everything we found when we
started adding in all these models that included
environmental impact that McDonald's and the amount of
meat that we are using, and the cows that we are
supporting, is such a huge impact on global warming that
we are going to go out of business just because
everyone's going to die.
And the way they actually did it was a little bit less fake
sounding than this.
And there were a whole bunch of blogs that
picked up on the story.
That this is really how people need to behave, that this is
how people have to be responsible.
That they'd quit and they put their jobs on the line and
their reputations.
There was a great deal of excitement about this very
very briefly until eventually it turned out
to have been a hoax.
So it's an example of the risk that you can make by being at
the forefront of the story in that it was also very very
embarrassing for a number of fairly high profile writers
who had taken this hook line and sinker.
So that's just one example.
But there's obviously fashions in music.
Interestingly enough, it's also one of these things
that's moving increasingly into that space between public
presentations of self, but what you're taste in music is
really puts you in a particular affiliation.
And it also is something where it requires being able to keep
up with what is the right thing to listen to.
So it certainly functions as a fashion.
Management is prone to fashion in a tremendous way.
20 years ago if you were a fashionable consultant, you
could talk about quality circles.
Today you talk about social networking.
But there's enormous amount of fashion in this information
based world.
Some of that may have to do with the actual value maybe
particular ways of doing management
are better than others.
But a tremendous amount of it can also be notions of fashion
of saying I am up to date within this world.
This is fashion in modern art.
This is painter John Currin who started
painting in this style.
He's now the top contemporary artist with some of the
highest sales.
He's had one man shows at all the big museums in New York.
When he started painting in the 1980s, there was very very
little figurative painting at all, and he and his peers
changed the style of painting.
But the whole contemporary arts world is very very
fashion based.
So these are just some examples of the different
dimensions, not just clothing, that fashion exists in.
But what's interesting here to think about in terms of
signaling and how you can think of this in terms of
design is that if you think of it as a form
of information prowess.
Some of the things that need to be balanced again here, is
this notion of the cost of the signal verses the benefits
that come from it.
So here again just showing some ubiquitous examples of
the types of fashion.
So let's look at something like an iPhone or little
bluetooth ear pieces for phones, they certainly are
functioning as fashion.
Almost any kind of technological innovation has a
big fashion side to it.

It may be programming languages ruby on rails can be
thought of as a fashion.

When you adopt, whether it's a new technology a new
programming language, when you come in at the very early end
you take that risk that you're going to do something that's
going to kind of fall off the world.
You may start to learn a new programming language that it
turns out after you've spent six months studying it,
learning it, becoming an expert in
it, it kind of folds.
Nobody ever uses it.
It becomes unsupported.
All your effort was in vain.
On the other hand, if you learn a new programming
language and you're the first one to do it and it has these
new capabilities, and then you get to tell all your other
colleagues about this great language you knew and then
they start to realize that everyone's using it.
But you're the one that they heard about this from first.
It gives you that status of being shown to be much more
knowledgeable to be at the forefront.
So when you look at something like this in terms of the
economics of signaling, there is the fashion signaling that
says where am I in this adoption curve?
And then you start to weigh it between the
benefits and the costs.
Because if you look at things as innovation, there's a
certain level of benefit that you get anyway.
And then as we see from signals, that benefit fights
against it as a signal.
Because the benefit could be if something is
obviously just better.
It doesn't function as well in some ways as a fashion signal
because it can be adopted by many people without being
particularly risky or showing their commitment to that
particular hierarchy.
Whereas if it has either some great difficulty associated
with it, and that difficulty whether it's a difficulty of
learning or it's a style of painting that to the
contemporary eye looks really ugly.
It has that risk to it that says I'm willing to pay the
cost of this risk in order to adopt it.
And I think what's also particularly useful about
thinking about it today is understanding the dynamics
within fashion both as the signal and how the benefits
and costs that are associated with innovation adoption
playoff it in terms of the signal.
Because a very interesting thing that we're seeing now
is, it's important in two things, one is that it's
accelerating at a fantastic rate.
Most accounts of fashion as a cultural phenomenon put it at
around the 15th century.
15th and 16th centuries is really the
beginning of fashion.
Certainly in clothing but in many other things.
Remember when we talked about when someone is copying a
signal, if you look at costume, for instance, in
medieval Europe for one people's clothing did not
change from generation to generation.
The things your parents were wearing around the time you're
born we're probably the things you would wear
throughout your life.
There were differences in costume based on class.
But for various reasons between simply very very
limited notions of social mobility plus sumptuary laws
that forbade copying things.
People did not say I'm going to achieve social mobility by
imitating the appearance and actions of those in a social
position I would like to have. It was completely unheard of.
As our society became more mobile, I'm not going to give
a big general history of western civilization of
course, but one way of understanding the role of
fashion is that as society became more mobile and the
possibility of trying to change social class through a
number of reasons some of which are changes in wealth
but also through imitations of particular kinds of behavior
was really the birth of fashion.
But in the 16th century, there's reports of princesses
in Poland writing to their advisers that they really
wanted to see dolls from Paris.
A lot of the fashions would be distributed through these
little dolls.
But it would take a year for the doll to get from Paris to
a country manor in the outskirts of Poland.
And so the rate of change in fashion was very slow.
Society with more static but also simply information moved
very slowly.
We're now living in an era where information moves across
the globe instantly.
And one of the things that's very interesting is whether
you're looking at it in terms of clothing, of music, of what
kinds of information, whether it's from blogs or news feeds,
that you need to be able to keep up with on a daily basis
in order to maintain a level of general knowledge that's
expected has rapidly accelerated.
And so understanding how people are using these
different ways of both accessing information but how
information itself is becoming a way of self presentation and
of identity I think is very very important.
I think that's one of the pieces that's particularly
relevant to the type of work that a lot of you, I think,
are doing here is understanding what it is that
people want to do with all this
information they're getting.
We're developing huge numbers of technology that say here's
a news feed, here's a visualization, here's a whole
new way of getting at this stuff, but this is about
looking at why do you look at things like Myspace pages that
have news feeds and blogs feeds and everything?
How are people using information as a way of
positioning themselves in an increasingly both mobile and
information based world?
I'm going to talk about one example of a piece that we
developed in response to this.
And I'm going to open this for questions.
This is a piece called, I'm going to talk about two
things, about a piece called urbanHermes.
And what we're interested in here was finding ways to take
the fast fashion of the online world into
the physical spaces.
And so what this is is it's a messenger bag.
Because we can forsee this as being something that was
clothing or anything else but right now, considering the
amount of technology we had to put in back of it, sticking it
into a messenger bag was fairly useful.
But it's a display that would have a picture on it that you
could change, but it works on Bluetooth.
And so it could also sense when people were in your
vicinity that were using the same system.
Anyone could see what they were displaying.
But depending on the types of communication channels you
chose to open up with others people who you had certain
things in common with could copy what you had on your bag
and vice versa.
And so what's interesting here was what we were, I'm not
going to get into the details here because we're running
short on time, but what you could do here was you could
both be able to copy things instantly but not have it be
done without your actual actions and
decision to do that.
Much as suddenly shoes don't teleport on to you.
But also you'd be able to see the provenance of things.
So an image that you picked up from somebody else, because
you happen to be in the same club with them, you could also
track it as it had moved throughout society.
But it was all based on face to face encounters in tracking
how these things moved.
So it's really about taking the speed of the internet
fashion and bringing it into physical spaces.
And anyone who is interested, I can show papers on that.
Then I wanted to just leave with one completely different
thought on the notion of signaling is the
question, are you human?
And this is, I think, finally one of the most interesting
questions that people are going to be dealing with on a
daily basis.
And again, it's very interesting
as a signaling question.
I have a student who has been doing a lot of research lately
in Myspace developing machine learning algorithms for trying
to understand what are the different human-like
components of particular pages.
Which are pages that are created by people to talk with
other people?
Which are much more corporate ones that are put together by,
whether it's band or increasingly companies
advertising something?
And which messages are being sent out by agents pretending
to be humans when they're not?
Is there some kind of pattern for ones that claim to be
human but in actuality are not human?
And this becomes a question that, I think, has tremendous
implications in the next few years.
That not only is this something that people are
going to be trying to understand, whether it's in
social networking sites where there is increasing numbers of
pages that are in actuality corporate entities or jokes,
or something else but are saying that they are human.
To chat spaces, all kinds of discussions where there may be
participants that are not, in fact, human.
But also on things where we have to ask underneath some of
the interesting questions about actually when
is it that we care?
There's, in Japan especially, increasing numbers of robotics
projects saying here we can build a robotic animal that
will be a companion for the elderly and the lonely.
And so what that it doesn't really have any feelings.
It can act so well that it has feelings it can
make you feel better.
And so the signaling questions that we've been talking about
here when put into trying to question what does it mean to
be able to signal you're human start to address
some of these issues.
Anyway, thank you.
And I'd like to welcome any questions.

AUDIENCE: [INAUDIBLE]
essentially it seems like that you are trying to change your
signal because you can't manifest your signals by a
very specific quality but when you project something on your
screen, some of your interests, and you have people
coming by looking at your screen with similar interests
you project a different personality essentially.
You can be a shy person then it can display your interests
on a little screen on your back.
And make other people contact you based on that.
So you change the signaling this way, right?
JUDITH DONATH: It really is much more in the tradition of
the fashion, because part of it is you can see where in a
whole trajectory were they?
So once you say I'm interested in something you
adopt it from someone.
What you're able to see is were they the
first or the second?
Are you at the very beginning of this or are you the 500th
person to take it.
So it tends to be more function
within this realm of--
and then you also get to see how far off it spreads.
So it's much more a signal of position within a
hierarchy of adoption.
So part of it is any kind of fashion resource is always
encoded in something else whether it's clothing or
management style et cetera.
So the content of it has that information that says if it's
a picture of Hello Kitty do you judge that this is
something you want to be adopting because of your
general aesthetics and knowledge and other things?
So it's using that content to give you a cue about where you
will end up there.
AUDIENCE: So the fashion statement is just a
demonstration on the surface, but then it causes people to
communicate more in their real life because they were already
used to assume different personalities in the online
world, right?
But they tend to be very shy around other people.
But now you can project your interests on your little
screen and have more communication with people with
similar interests.
JUDITH DONATH: Yeah I didn't get into that particular issue
here but a tremendous number of the projects that we do
that are in our work as a whole are about trying to find
ways to break down a lot of the barriers between people's
communication in a face to face basis.
Obviously, that it's very very open when you're at a distance
from others, but how can we use these kind of hybrid
spaces to break down the barriers?
Yes?
AUDIENCE: Can you examine the adoption rates for fashions
when signals are signal that has a high cost in one area so
maybe it takes time to learn it or to go out and order the
product that you but the actual cost in
monetary terms is low.
So instead of Apple introducing an iPhone at
signaling not only that it's costly to learn this new
system but also costly in monetary terms. But they
introduce it at $200 how does that affect adoption rates?
JUDITH DONATH: Yeah.
There's a lot of research in that area.
There's a book called The Diffusion of Innovations by
Everett Rogers who did a lot of work
on innovation research.
And what I've done with the signaling theory is there's a
long history of theories about fashion that are based on
Georg Simmel's work on imitation and differentiation.
It basically says those who are of lower on some hierarchy
are trying to imitate those who are above them and those
above them differentiate.
The innovation research looks at what turned out to similar
phenomenon but comes up with a very different model which is
that people want to be in different places, in some kind
of adoption.
And he has this notion that there are early adopters and
later ones and people want to be in these particular spaces.
And one of the contributions of looking at it as a signal
is it's a way that you can take both of these models and
unite them to show that the costs and benefits of the
innovation can function as the costs that ensure the
reliability of the signal in terms of social signaling.
Because the innovation work tends to really ignore how
much people do things for social status.
He has one little footnote saying oops I should talk
about this because we really don't, but in our interviews
people always deny that they're
doing things for status.
Very very few people, no matter how trendy the thing
they have just gone out and done or said or bought, says
yes I was doing that to look like I'm even more high
status than I am.
They will always come up with some reason why it was
practical, or useful, or a tremendous sale.
But, they will almost never admit it so if your research
is all on looking at why people do that, but the price
point thing is fairly common in marketing and research.
And here it's more about trying to find a unified way
of bringing in the social signaling aspect that, I
believe, is a tremendous driving force that is much
often ignored in most of the marketing world.
So yes?
AUDIENCE: I was wondering if you could comment on the
disruptiveness of computers in this environment.
Because a lot of your examples come from evolutionary
biology, where you have the peacock's tail or the
butterfly mimics something that tastes bad and that's a
very slow moving process kind of like
fashion in the 15th century.
But now instead of going out on the quad and saying hey can
I borrow a quarter from everybody I meet I can harvest
a million emails and send them out.
And the orders of magnitude of change is really dramatic, and
I must think that has an effect.
JUDITH DONATH: Yeah.
I think it has a tremendous effect and it's in some way
sort of the point of what I'm saying is that we are living
in an era where the rate of change of meaning is so rapid
that it becomes a tremendous cost in and of itself.
That time and attention and the ability to process
information has become a enormous resource that people
need to be displaying their--
it's that notion of information prowess.
So forget about evolutionary time, which is extremely slow,
but if you lived 700 or 800 years ago there certainly was
a lot of information you might need to deal with in your
daily life but it changed very very slowly.
So that ability to harvest new information wasn't something
that was a tremendous cut off in terms of resources.
But for contemporary people, the ability to both show your
ability to find the right information in a vast overload
of information, how you display that
prowess of doing it.
I didn't get into some of the more complexities of the
models that I have been developing here, because one
of the things that's very interesting about thinking
about how people display their ability their prowess with
information is that unlike other resources where you can
waste them and then they're gone information spreads.
It's not something that's
transferred from one to another.
You cannot throw it away.
The fact that I've told you something doesn't mean that I
don't know it any more as opposed to if
I give it to somebody.
And so I think a lot of this is in the work of trying to
understand what are the dynamics of information as a
resource as it moves faster and faster.

Any other questions?
Yes?
AUDIENCE: I was wondering about your slide where you
talked about the acceleration in fashion.
And I was just wondering what you actually meant by
acceleration.
So I can understand in the 15th century when fashions
need to travel from, let's say, Paris to England.
Sure, it's going to take like a month or something like that
for that to happen, but that's just one way of
looking at it, right?
It's one metric of saying well this is how long it literally
takes information to propagate.
But whether or not that name will actually catch on if the
English court at that time happens to be particularly
hostile towards the French throne then well it doesn't
matter how long those names take to catch on.
I mean it doesn't matter how long those names take to
propagate they're not going to catch on.
So similarly, now with the spread of information you have
information is diffusing much faster but also you're going
to have much smaller groups.
The number of--
assuming in the 15th century or 16th century you have a lot
more people in England who were willing to dress the way
the court is, right now that number's
going to be much smaller.
You have much smaller groups, and there's going to be
resistance to information influencing different groups
as it travels.

JUDITH DONATH: There's a couple of really interesting
points you've raised.
So let me just try and answer you while I can still keep
track of them.
So the last thing you talked about is really about the
fragmentation of society.
And when we talk about status and affiliation, a lot of
that's the affiliated aspect of it.
It's to say that the traditional notions of
fashion, for instance when Simmel first talked about
imitation differentiation, his model was of a single, very
stratified society.
I think you can look here in any number of multiple
subcategories and people exist within a
whole variety of these.
So there are fashions in music but people can participate in
different subcultures there.
There are many many different possible groups.
You can think of the nations as different ones.
There's subcultures within academics, not only within
particular departments, but people will move according to
a different hierarchy.
We can say there's a nearly infinite number of
hierarchies.
This phenomenon happens both within them and between them.
Your first question about what did I mean by the rate of
acceleration, I simply meant that the amount of time that a
particular form lasts as a signal of something being at a
particular point in some hierarchy has gotten much
faster the rate of change.
And if you're looking for specific examples there's a
lot of work, for instance, in clothing fashion.
Just in how much less time even in the last 20 years
particular things have stayed in fashion.
There's quotes from Diana Vreeland and [? Wintour ?]
talking about you know how the season in which they measure
what is a particular style has become much faster.
I don't have it with me, but there's an interesting graphic
of how long song's stay as hit music from the
1950s to the present.
And so I think it starts out as something just generally
staying as a top hit for four weeks, and now it's often down
to a period of days.
So it's simply that rate is changing, and I think a lot of
the cause is that information just moves much more quickly.
And so it's no longer a matter of having to wait.
If you want to listen to music, of having to wait until
your local record store, somewhere in South Dakota,
gets around to getting some music shipped out to them
before you can have that record.
If everyone can download it instantly, it's going to
affect that rate of change.
So I don't know if I missed some other main point.
Yes?
AUDIENCE: [INAUDIBLE]
to the information because that way I can say well I'm in
this group, and if this information doesn't apply to
me I don't have to take in all this giant flow.
If you're not wearing black and I'm a goth then I don't
have to worry about you.
Or if it's not Hip hop I don't have to listen to it.
And so I can ignore that information.
JUDITH DONATH: I think that probably is one of the reasons
for that type of fragmentation.

There's many other theories about why we live in a more
fragmented society.
I think some of it it's very hard to look at these things
without the historical perspective, because things
that look vastly different to us when we're living in the
midst of them may all just look like one big blur of
early 21st century things 200 years from now.
So I think some of it is also that the closer you are to a
particular era the more nuances you
see within the culture.
I think it's also very easy to look at history and see it as
a single unified culture.
But I think some of it has to do with how stratified a
particular society is and how much of a single common
culture there is.
So one more question, I think it's time.
OK.