Inkwell: Authors and Artists

    
Our next guest, Steven Johnson, is the author of _Emergence: The Connected 
Lives Of Ants, Brains, Cities, and Software_, acclaimed as one of the best 
books of 2001 by Esquire, The Village Voice, Amazon.com, and Discover 
Magazine.  Johnson was also cofounder and editor-in-chief of FEED, the 
pioneering online magazine, as well as a co-creator of the 
Webby-award-winning community site, Plastic.com. He was named by Newsweek 
as one of the "50 People Who Matter Most on the Internet." Johnson's 
writing has appeared in The New Yorker, Wired, Lingua Franca, Harper's, 
and the London Guardian, as well as on the op-ed pages of The New York 
Times and The Wall Street Journal.  The New York Times' Michiko Kakutani 
called his first book, _Interface Culture_,"one of the most thoughtful, 
literate studies yet published on the cultural impact  of recent 
technological changes." Johnson has degrees in Semiotics and English 
Literature from Brown and Columbia Universities. He lives in Manhattan's 
West Village with his wife and son.

_Emergence_ is what happens when an interconnected system of relatively 
simple elements self-organizes to form more intelligent, more adaptive 
higher-level behavior.  Drawing upon evolutionary theory, urban studies, 
neuroscience, and computer games, _Emergence_ is a guidebook to one of the 
key components of twenty-first-century culture. Until recently, Johnson 
explains, the disparate philosophers of emergence have worked to interpret 
the world. But today they are starting to change it. This book is the 
riveting story of that change and what it means for the future. 

Leading the discussion is recent inkwell.vue guest Derek M. Powazek, who 
is a writer and designer living in San Francisco. His book on the design 
of community spaces online, "Design for Community," was published last 
year by New Riders. Derek lives online at powazek.com.

Please join me in welcoming Steven and Derek to inkwell.vue!
  

    
Thanks, Linda!

Steven, let's begin by talking about how this book came to be. What
inspired/motivated you to write it? 
  

    
There's a funny story about the inspiration for this book. After I
finished writing the last one, I spent about six months doing very
early-stage research, trying to figure out where to go next, and I was
basically following two distinct paths: reading books of urban theory
and history, and reading books about brain science. I thought for a
long time that I was going to have to make an either-or decision at
some point, either a book about cities, or a book about minds. 

And then I happened to get as a birthday present a book of old city
maps from the mid-19th-century that included a map of Hamburg that
looked uncannily like a profile view of the human brain. And somehow
that image set off a chain of connections in my head, and I started to
think that maybe that there was a book to be written that was about
cities *and* brains. It ended up being more than that, of course
(there's more about ants than about brains, for instance) but that map
was the original spark.

Once it became clearer to me that I was writing a book about bottom-up
or self-organizing systems, I found myself drawing on a number of
earlier books for inspiration, books that had also wrestled with
complexity theory and emergence with a popular audience in mind. (One
of those books, of course, is the masterful Out of Control, by the
Well's own Kevin Kelly.) One of the things that I was trying to add to
the already impressive library was a more developed connection between
cities and emergence, a connection that appears here and there in the
literature of complexity, though it's rarely more than a passing
reference. So I wanted to bring urban development more fully into the
world of emergence. (Fortunately, Jane Jacobs had already done most of
the work for me.) 

And of course I wanted to connect all of these ideas to what had been
happening in the digital world -- on the Web, and in videogame culture
-- over the past few years. But I'm sure we'll have plenty of time to
get into that over the next few days...

This should be fun! 
  

    
Before we get into the content of the book, let's get our terminology
straight. Two words that appear over and over in your book are
"emergence" and "adaptive." If you would, please give us the basic
outline of what they mean and how you use them.
  

    
The emergent systems that I talk about in the book are systems that
are made of many lower-level constituent parts, each of which follows
relatively simple rules of interaction and lacks an awareness of the
overall state of the system. Out of the semi-random exchanges of these
many agents, a higher level order arises: ants organize into colonies,
urban dwellers into neighborhoods. That movement from low-level
interaction to higher-level order is what we call emergence. 

"Adaptive" is a key term as well -- the systems I talk about are often
called "complex adaptive systems." I stressed the term quite a bit
because the systems that I'm interested in aren't just examples of
patterns emerging out of seemingly random interactions; they're often
patterns that are *good* for something. The emergent behavior of ant
colonies helps them pull off incredible feats of resource management
and engineering; neighborhood formation helps cities organize and store
collective information and makes them more intelligible spaces.
Sometimes that adaptive behavior is the result of an evolutionary
process  (as in the ant colonies); sometimes it's the result of direct
human interaction (like some of the software programs I look at.) But
in all the adaptive systems there's some feedback mechanism pushing the
system towards a more efficient state... 

I'd add one thing here: the book is not exclusively focused on
textbook examples of complex adaptive systems. There's a kind of
philosophical approach to these systems that's usually called
"bottom-up" as opposed to top-down hierarchies. So I was interested
generally in systems that didn't need leaders, systems that organized
from below. And that led me to write about things like the Amazon
recommendation engine, or the anti-WTO protest movements, that
technically aren't complex adaptive systems, but nonetheless share a
certain sensibility and organizational structure. 
  

    
Hi Steve! Congratulations on a wonderful book. I loved it.

Are you acquainted with Anthony Townsend at NYU? He has some interesting 
things to say to city planners in regard to the effects of wireless 
communications.

Have you thought about emergent behaviors of intelligent units (i.e., 
humans) who are able to "swarm" in new ways through the use of mobile 
communication technology? Ronfeldt and Arquilla talk about the swarming 
tactics used by WTO protestors in Seattle (in the book "Netwar), and I 
thought about possible connections with the emergent behavior of dumb 
units that you, Kevin, et.al., discuss.
  

    
Welcome, Howard! Say ... aren't *you* writing a book a bit like that?
;-)

Steven may be limited to a once-a-day check in here at the beginning,
so it's okay if we double up on questions a bit. Everyone should feel
free to post their thoughts! That said, I have another question.

Steven, one of the things that struck me while reading your book is
the whiplash pace you hop through diverse examples when discussing an
idea. In just a few pages you'll go from discussing the media, to air
conditioning systems, to the way our brains work, to the way computers
work, to the web ... and it's all tied together. 

I really like this kind of associative thinking, and I know I've
always thought this way. But from my experience in conversations,
sometimes I bump into a linear thinker who just won't follow me from
hop to hop.

So I was wondering, did you have any editors or reviewers just didn't
*get* it when you were writing? Was it hard to communicate the vision
for the book in the beginning? Have you heard from anyone who is just
lost by it now?
  

    
Hi Howard, very pleased to hear that you enjoyed the book. I don't
know Townsend or his work. What department is he in? (I'm actually
teaching a grad seminar on "Emergent Communities" at NYU this
semester.) The wireless urban swarming idea is a fascinating one --
I've been thinking a little about what happens when we have a full
markup language for GPS co-ordinates and people starting filling the
real world with virtual annotations. Eventually you might start seeing
the kind of taste-clustering that Amazon does enacted in urban
environments -- imagine exploring a new city that way: "people who
liked this block also liked these other blocks." Or even better:
"people who liked some other block in some other city also liked this
block in this city."

Derek, as far as the whiplash jumps go: I had more trouble with that
in the last book, where I think a lot of people picked it up expecting
it to be a straightforward survey of interface design, and suddenly
found themselves reading about Dickens. Emergence makes it clearer from
the outset that it's interested in a pattern of behavior that exists
in many different disciplines and at different scales. (The subtitle
alone prepares you for the jumps.) So I think people have had an easier
time following the threads this time around. 

The one audience for whom the pace can perhaps be a problem is serious
complexity theory hounds who've read all the popular literature and
dabbled in the more academic stuff; I think they can sometimes wish
that I'd gone into more detail on, say, the ant colony material --
spent thirty pages on it instead of ten.  But I think that would have
been a different kind of book. There have also been a few folks who
object to the connections made to the videogame world -- it's almost
like there's something demeaning in talking about SimCity in the
rarified context of complexity and self-organization. But I suspect
these are largely folks who haven't actually played SimCity yet... :)
  

    

Notions like emergence and self-organization have fed into political ideas
about devoluting the state, which are really problematic.  Do you deal with
that in the book and if so, how?
  

    
RU, I would say that I deal with the political implications of
emergence gingerly in the book -- it's such a thorny issue, and I think
you're right that theories of self-organization have a history of
being yoked to anti-government positions. I was worried that if I spent
too much time on the politics topic it would turn into a distraction.
So I dealt with it in two primary ways: I tried to make the point, at
several moments in the book, that bottom-up approaches were not a
cure-all, and weren't necessarily intrinsically better than top-down
ones. Cities, for instance, solve a lot problems from below, but they
still need top-down, state-subsidized fire departments, etc. 

And then I also tried to make that point that if you were set on
applying complexity theory to political systems, and eliminating all
the top-down forces in a society, then the state was only one potential
target: the modern multinational corporation is probably one of the
most top-down, hierarchical structures on the planet.  If you really
believe in the power of self-organization, then the big companies have
to wither away alongside the big governments (at least the governments
these days are increasingly chosen by bottom-up democratic systems.) To
that end, I included a brief nod to the anti-WTO protest movements at
the very close of the book, talking a little about how they had modeled
their political structure after swarm systems... 

Apparently, I'm told that the book is developing a little bit of a
following the quasi-anarchist, protest community -- when I've done
radio call-in shows, the enthusiastic calls that have a political slant
tend to be from anti-globalization folks, and not hardcore
libertarians. 
  

    
Well, that would lead me to ask, where does "self consciousness" enter
in.  

You talk about systems where the constituent parts "lack an awareness
of the overall state of the system."  Does it make a difference if
those parts *want* to have an awareness of the overall state?  

In some ways, right now, it seems to me that the more sophisticated
the constituent parts are about comprehending points to intervene in
the system (the grander their assumptions about what they think they
know about the overall state) the more likely they are to make
"mistakes" in their actions.  Like how urban redevelopment screws up
neighborhood self-organization more often than it furthers it.  

And yet, for anti-globalization folks, looking for points to intervene
in the system feels like the only game in town.
  

    
Some comments:

> So I dealt with it in two primary ways: I tried to make the point, at
> several moments in the book, that bottom-up approaches were not a
> cure-all, and weren't necessarily intrinsically better than top-down
> ones. 

Good point, and it raises a question for me... how you apply normative 
thinking to (or within) self-organizing systems. It seems to me that 
practicality would significantly outweigh moral and ethical principle in 
guiding their evolution. In fact, don't we see that in 'realpolitik'?

 
> the modern multinational corporation is probably one of the
> most top-down, hierarchical structures on the planet.  If you really
> believe in the power of self-organization, then the big companies have
> to wither away alongside the big governments (at least the governments
> these days are increasingly chosen by bottom-up democratic systems.) To
> that end, I included a brief nod to the anti-WTO protest movements at
> the very close of the book, talking a little about how they had modeled
> their political structure after swarm systems... 
 
But I'm missing how you connect the dots here... e.g. how does the 
self-organizing "nodal politics" of the protest movements have any actual 
impact on corporate organization/behavior? Won't this be driven more by 
business revolution (which is also sort of in the air) than political 
wonking and maneuvering?

(It's interesting to me to see the, er,  emergence of limited liability 
companies as the busines structure of choice these days... they tend to be 
flatter and more flexible than corporations... but that's a digression).
  

    
The swarming tactics of the WTO protestors were effective street tactics, 
using small, autonomous but strategically aligned groups, who 
independently acted under a previously agreed general plan, coordinated by 
real-time communications using mobile phones, Internet sites, and pagers. 
At least as described in Ronfeldt and Arquilla. That's not the same as a 
political philosophy.
  

    
Great stuff, all! 

Steven, I know you may want to respond to the above stuff (I'd
especially like to hear what you have to say to keta's comment #10),
but I wanted to throw another log on the fire for when you log back in.

When we spoke a year ago for Design for Community
(http://designforcommunity.com/display.cgi/200109181843), you were
doing a lot of thinking about FEED and Plastic. I was wondering: How
did your experience with content-based virtual communities inform
Emergence? And, in hindsight, with FEED on ice and Plastic struggling
back to life, what have you learned about emergence on the web? Did
either system ever get there, in your eyes? 
  

    
I've had a day of travel to think over Keta's question about
self-awareness, which I read just before heading to the airport. It's a
great question. As Howard rightly points out, there's a simpler
answer, which is to say that the protest groups were "swarming" in
Seattle in physical ways, in terms of how they choreographed their
movements through the streets. But let's take the question on Keta's
terms, in terms of broader political strategy and not just protest
strategy. 

It seems to me that the question of awareness revolves around what
motivates your actions, and not just your overall assessment of the
global system. Think about it in city terms: we're aware of the system
of neighborhoods, but the individual decisions out of which
neighborhoods are created -- to visit this restaurant and not this one,
to cross town to visit this market, to move closer to your kid's
school -- are motivated by forces one level down the chain. You're
thinking *and* acting locally, and yet the sum of all those decisions
creates the global order of neighborhoods. You don't patronize a
certain bar because you're actively trying to create a certain kind of
neighborhood. The people who are actively trying to create a certain
kind of neighborhood tend to be the planners, who nine times out of ten
end up screwing it up. 

How does that translate into politics? The other swarm-like property
of the Seattle protestors was the fact that they were a gathering of
loosely affiliated interest groups, with no common, overarching
ideology. (And no clearly defined leaders.) They were certainly aware
of the broader system that they were trying to insert themselves into,
but they were motivated by a diverse group of local causes. It's a kind
of mixed-use version of protest politics. What I like about that
approach it how it mirrors one of the core principles of emergent
systems, which is their dependence on random encounters between agents.
Ant colonies and sidewalks both rely on the random swerve of an ant
stumbling across the pheromone trail of a neighbor, or a city dweller
stumbling across a new shop on the way to somewhere else. The
intelligence of the overall system relies on that randomness -- it
can't grow or evolve without it. So there's something fitting in the
protest groups clustering together in interesting new configurations,
each cluster made up of different ideological perspectives colliding
with one another in new ways. 

I suppose it's a rainbow coalition in new clothing, but there's
something more deliberately chaotic in the Seattle model, and less
touchy-feely... 

JonL, nice to see you here. Can you rephrase the normative question?
I'm in South Miami Beach, so perhaps the sun is slowing down my brain,
but I didn't quite understand it...
  

    
Heh, let's see if I can remember what *I* was thinking!

I'm just wondering about the ethical dimension of 'swarms' or 'hive mind' 
or whatever we call emergent entities, especially the activist adhocracies 
mentioned above. It seems to me that emergent entities are inherently 
driven by the practical, and not so much by higher principles. And I'm 
thinking that some of the best moments in human history occur when 
principle is given more weight than pragmatic thinking. So I'm thinking 
that we might lose something when we organize from the bottom up, and 
emergent forces drive behavior and, presumably, evolution.

It was just a thought, though... I've got peripheral thoughts bouncing 
'round that contradict my comment... e.g. maybe the emergent behaviors 
can be driven by principle just as well as by practical concerns.

I've been back and forth on this. I've preached the promise of cyberspace
adhocracies, but at the same time I'm thinking that an angry mob may be an
emergent entity, too...
  

    
why couldn't altruism be an emergent property?  I don't see the obstacle.
  

    

Al Qaeda of course also fits to this model, or a close relative of it...
which tells us the obvious, that there's no intrinsic moral value to its use
as a tactic, even if there's an implicit anti-authoritarian idea behind it
in the abstract.
  

    
 I don't think that it's certain the al-Qaeda fits this model.  That
was the conventional wisdom a few months ago, but it seems like they
had a more top down structure than was originally suspected.
  

    
Arquilla and Ronfeldt's anthology gets into some detail about terrorist 
social networks.
  

    
I think RU and I are close in our thinking on this.

Pete: I think a "group mind" or swarm or whatever would tend to be 
pragmatic - altruistic if it fits the group goal. But my question was 
really about situations where the principled behavior would contradict the 
more pragmatic goal or solution. 
  

    
Jon, thanks for the rephrase. Let me try to answer in a way that
potentially connects to Derek's question about online communities. I
talk a lot in the book about the Slashdot/Plastic model of group
filtering and feedback -- where contributions are rated by other users,
and well-regarded contributions result in "Karma" points doled out to
users, which give them special privileges in the community. I talk
about this as a bottom-up approach to editing, as opposed to a top-down
approach where you hire a bunch of full-time editors who comb through
the site and delete obnoxious posts and promote the insightful ones. 

Now, there is a danger in this kind of self-organizing approach, which
is the "mob rule" idea that you allude to: the ideas that get rewarded
are those that most Plastic/Slashdot users already agree with; ideas
that are out of the mainstream could potentially be shut down by the
group consensus. (This is a real issue, since you can filter out posts
that are below a certain quality threshold on both sites.) This has
produced a number of fascinating -- though sometimes distracting --
arguments on Plastic: because it has effectively become a politics
weblog, unlike Slashdot, there's a constant debate over whether 1) the
site slants to the left in its politics, and 2) right-leaning posters
are given poor ratings by the community just for their political views.


So does that mean that self-organizing community systems necessarily
sacrifice the "higher principle" of diversity? (Leading to a kind of
collective Daily Me.) I don't think so. One of the points I make in the
book is that the Slashdot/Plastic model may trend towards a groupthink
-- though it practice I wouldn't say that it has resulted in too
narrow a focus -- but that trend is only there because of the specific
way that the software has been written. You could just as easily create
a bottom-up system that would promote diverse voices (and controversy)
simply by tweaking the rules a bit: let everybody rate each post, but
instead of promoting the highly-regarded posts, you promote the posts
that have attracted the strongest responses, both positive and
negative. In that kind of system, the fringes sudden come alive:
instead of favoring the center, the Rush Limbaughs and the Michael
Moores suddenly rise the top -- the voices that you either love or
hate. 

In that kind of system, you'd have a local motivation of "do I like
this guy or not" creating a higher principle of diversity. You just
need to set up to the rules properly to make it work. 
  

    
Good idea - do you know anyone who's doing it?
  

    
Jonl - Anyone *could* do it, just by tweaking the Slash code that runs
both sites. An interesting idea.

Even better: Allow each user to pick, or have some kind of "wizard"
that asks the user a series of questions, and then intuits what kind of
filter he/she would best respond to.

But my question is still this: Are these systems necessary to create
emergence on the web, or are they merely icing? In other words, are we
really seeing emergence on the web yet? Or does the slime mold still
have us beat?

Personally, I do think there are emergent patterns in the web, but
they are very hard to measure in terms of one site. Because the net is
so diverse, with so many users doing so many things, in order to look
for emergent behavior, you have to look at what people do in total.
Blogdex (http://blogdex.media.mit.edu/) comes close, for example, but
that only looks at links (not email or IM or any of the other ways
people communicate online).

And just to echo some of the other lingering questions: Is an angry
mob an emergent system? Is altruism an emergent property? What about Al
Qaeda? When you get the chance, Steven, I'd love to hear your
thoughts.
  

    
I think we had "emergent" political forces within early net.activism, some 
more successful than others. Shabbir Safdar and Jonah Seiger told me that 
their "black web page" campaign in response to the CDA couldn't happen 
today, because the web is part of an industry... then it was an 
"electronic frontier," and you could aggregate meaningful support by 
unleashing memes. You have to have the right kind of petri dish to 
colonize dissent, I guess.
  

    

My personal experience is that anything I try to grow organically virtually
doesn't even sprout, if it has a bit of money and gets mainstream publicity
it gets to live awhile.  Everybody always tells me the counterculture model
should work for me, but it never has. But maybe that's just me...