Some notes on the importance of overlapping knowledge, the supposed
democratizing effects of the Internet, the supposed decline of B2B,
and the supposed conflict between teaching and research.

By popular demand I've put titles on the longer notes.  I'm reticent
to do this because each set of notes is designed to fit together as
a whole.  But I guess people have lives and don't have time to read
straight through to find the most relevant bits.  So, okay, titles.


I've added more useful stuff to "Networking on the Network".  The most
important change is a major rewrite of the section on negotiating an
academic job offer:

The previous version had been dangerously unclear on the distinction
between a department wanting to hire you and a university making a
legally binding offer.  I haven't been on a search committee in a few
years, though, so I'm hoping that faculty can look at the new version
and tell me if I've got it right.

Have you heard any horror stories from PhD students or junior faculty
members that might suggest further topics for "Networking on the Network"?
If in doubt, send them along.  I've already heard some killers, but I'm
sure I haven't heard them all.


Longtimers will also recall the obsessionally exhaustive bibliography
of "books on the social aspects of computing, 1996-1997" that's at:

Well, I've created an analogous bibliography for the years 1994-1995.
It's not nearly as obsessional or as exhaustive, but I think it's
useful to look at the development of what was being published when.
(Lissen, George W. Bush plays solitaire; I assemble book lists.
Sometimes a guy's gotta zone out.)

At some point I'll merge the two.


People often ask me to recommend books that students and citizens
can use to understand the social role of information technology from
a realistic perspective, freed from the hyperbole that everyone is so
tired of.  Here are the books I recommend:

  Rob Kling, ed, Computerization and Controversy: Value Conflicts and
  Social Choices, second edition, Academic Press, 1996.

Though it is a few years old, this volume is infinitely preferable to
the transient hype.  It's a diverse collection of articles on the full
range of social issues around computing: privacy, online interaction,
safety, ethics, the design process, organizational changes, politics,
educational applications, and so on.  I used it as the primary text in
a required course on social issues for computer science undergraduates
a few years ago, and it worked perfectly well.  Fifteen years ago such
a course would have been a disaster; the students would have figured,
"instead of taking this course we could be learning more technical
stuff, therefore it is a waste of time", and mass resistance would
have broken out.  But now I think that people are more accustomed to
the idea that computing is a social activity.  The cultural impact
of the Internet gets much of the credit for this.  I did have a
mass exodus of people who were terrified of the writing requirement,
and I also had one of those guys who give fraternities a bad name.
But mostly I was able to sell the material.  I said, "if you want
your technical work to get used then you need to understand the
context in which people will be using it, or refusing to use it".
That seemed to do it.

  William Dutton, Society on the Line: Information Politics in the
  Digital Age, Oxford University Press, 1999.

This is a synthesis of a large-scale research program in the UK, the
Programme on Information and Communications Technologies (PICT), from
the mid-1990s.  Even though Bill is from USC, they brought him in as
director during the later stages of the project, and he put the whole
thing together.  I probably tend to underestimate this book because
I know all the research on which it is based -- including Bill's own
work over many years.  But if I step back and look at it through the
eyes of someone who is not acquainted with serious, methodologically
sound empirical investigation of the reality of computing, as opposed
to the fairy tales of the enthusiasts, I can see what an incredible
gulf he is bridging.  His organizing theme is "shaping access": access
of people to information, of audiences to programming, of businesses
to their customers, of citizens to the government, and so on.  Because
computing is so malleable, we cannot know a priori what "impact" it
will have.  Instead, we have to look at the "ecology of games" --
people politicking and strategizing in various economic, political,
and cultural venues -- through which the practicalities of access get
shaped in the real world.  Despite its academic grounding, the book
is written in an accessible way, and it would make a good textbook for
serious students at any level, not to mention regular citizens who are
looking to vent the hyperbole from their heads.


The importance of overlapping knowledge.

On one level, an institution is a set of roles and a set of rules.
That's the *formal* level on which we all get defined as doctors,
patients, teachers, students, defendants, jurors, coaches, players,
and audience members in our various dealings with one another.  On
another level, an institution is a body of knowledge.  The people
who occupy those roles and confront those rules develop a body of
intuition, of lore, of savoir-faire, of settled practice, of maxims
and how-to's.  That's the *substantive* level on which the collective
learning of society gets applied to practical outcomes for better
or worse.  Here are some examples of the substantive level:

 * The detailed manufacturing knowledge that is accumulated by
   the engineers in an industry (a phenomenon first identified by
   Thorstein Veblen and recently expanded upon in the work that Alfred
   Chandler summarizes in the passages I quoted the other day)

 * The highly evolved strategies for networking and career-building
   in the research community that I have outlined in "Networking on
   the Network".

 * The skills and customs for collective problem-solving that build
   up in the political culture of a democracy (see, for example, Harry
   Eckstein, Division and Cohesion in Democracy: A Study of Norway,
   Princeton University Press, 1966).

The two sides of an institution, formal and substantive, presuppose
one another and play off against one another.  Both are part of a
big story about human activities get coordinated, and how people
manage to be so brilliant collectively even though they are so finite
in isolation.  Both sides of the institutional story can become
invisible, taken for granted, because the framework of institutions
that organizes a society does not often change.  So it is easy to
adopt simplistic or even destructive attitudes toward institutions.
Proponents of technology-driven change look at the gathered wisdom
of institutions and see pure reactionary resistance to the imperatives
of progress.  Conservatives, at least when it benefits them to do
so, claim that only a blind reverence for tradition will enable the
accumulated wisdom of institutions to persist and society to avoid
falling into chaos.  That is what conservatism means.  Realists occupy
the rational middle ground between these two extremes.  They realize
that the accumulated knowledge of institutions is both valuable and
a hindrance.  No institution is perfectly just or perfectly efficient,
and much of the settled practice of any institution consists of hidden
interests and routinized log-rolling.  We cannot go around randomly
blowing up institutions, because we do not know how to fabricate new
institutions to replace them on short order.  Yet we cannot simply let
them be, since the world can most assuredly be better than it is now.

Because institutions exist largely to solve informational problems,
these things matter especially in our current times of radical change
in information technology.  Yet we know so little about them that a
great intellectual vacuum is opening up.  The fall of communism and
the spread of democracy created intellectual interest in the nature
of institutions.  Economic analysis of industry structure is proving
a valuable tool in parsing the unexpectedly complicated patterns
of change that new information technologies are bringing to markets.
And historical studies have shown how institutional take form through
disputes between social groups.  Even so, we still know almost nothing
about institutions and the ways they change.

So in reading the literature on the subject, I try to articulate useful
intuitions about institutions -- intuitive ways of explaining how and
why the social world works -- so that we can have a chance of moving
toward fairer, healthier, and more efficient institutions as the
possibilities afforded by new information technologies begin to unfold.
I want to sketch one of these intuitions, which I'll summarize using
two concepts: anamorphism and overlap.

First, anamorphism.  You've seen Saul Steinberg's cartoon, "View of
the World from 9th Avenue", that was on the cover of the New Yorker in
1976.  It's a map of the world, but with lots more detail in Manhattan
than anywhere else.  The further away you get from 9th Avenue, the
less detail, until the American West is just a cactus and Japan is
just a blob on the horizon.  In mathematical terms Steinberg's map
is anamorphic: relationships of geographic locality are more or less
preserved -- stuff that's close together in the real world is close
together on the map -- but the map is grossly deformed, as in a
fun-house mirror, so that some parts are much bigger than they ought
to be, and other parts are much smaller.  The joke was on New Yorkers,
but everyone knows that the lesson applies to them as well.  We all
know our corner of the world the best, and none of us knows the world
as a whole.  Parochial or not, it's simply impossible to know the
whole world.  We're all locals.  But our knowledge is not limited to
our immediate vicinity.  Every one of us has some vicarious knowledge
of many other parts of the world: through our past experiences, our
friends and family, people we meet socially, the newspaper, novels,
movies, etc.  The knowledge is often sketchy and distorted, but it
doesn't vanish at the end of the block.  It's an anamorphic map of
the geographic world and of the various social, professional, and
cultural worlds, including the ones we inhabit and the ones we don't.

Next, overlap.  You have an anamorphic map of the world, and so do I.
Your map is centered in your home and neighborhood, your office and
profession, your social network, your reading material, your resume,
and so on.  My map is centered differently.  When I travel, I often
ask myself, "What's it like for this, here, to be the center of one's
world?"  I chat with the elderly couple who run the dim sum shop on
the side street in Chinatown, and that's the center of their world.
They know the regulars in the shop, the politics of small business
people in Chinatown, their relatives, the news from China, their kids
at college, and so on.  Of course, they also know about US national
politics and the Internet and everything else, just like anyone else
does, just like I do.  It's just that the proportions are different.
Their map is deformed; so is mine.  They have a 9th-Avenue knowledge
of things that are like the lone cactus in the West for me, like how
on earth people manage to stay married for fifty years.  I probably
know a few things really well that they've spent maybe ten seconds
thinking about.  The point is, our maps overlap.  We know many things
in common.  We *can* chat because we have a reservoir of references
that we can make in common.  We don't live in different worlds -- we
live in the same world.  We just have different anamorphic maps of it.

Anamorphism is a measure of our finitude and difference.  Overlap is
a measure of our universality and commonality.  The relations between
our maps are not random, but neither are they especially predictable.
And institutions depend on anamorphism and overlap.  Take the case of
the institutions of research.  The idea of research is that everyone
is supposed to do something new all the time.  It's very hard to do
anything new.  And it's hard to run an institution in which everyone
is always doing something new, because the institution won't work
unless it can evaluate the work, allocate resources, and create the
right incentives.  The need to credit all relevant work motivates
everyone to develop an extensive map of the literature.  The sheer
magnitude of the literature ensures that these maps will be anamorphic,
since nobody could ever read it all.  The need to differentiate one's
work from everyone else's means that everyone's anamorphic map will be
centered in a different place than everyone else's.  But the maps will
overlap a great deal.  At least, everybody's map will overlap a great
deal with many other people's maps.  Junior scholars typically have
very focused maps; senior scholars typically have more extensive maps.
The senior scholars, being older, have had more time to map things,
but one's role also shifts with seniority, so that one is called on
to set agendas and evaluate work that encompasses larger territories
beyond one's immediate speciality.

Anamorphism and overlap work together to keep the institutions of
research reasonably healthy.  Peer review means that everyone's
work is evaluated, and feedback on it is generated, by people whose
maps overlap enough to evaluate it responsibly, but whose maps are
nonetheless different enough that fresh perspectives are brought
to bear.  So I might write about the role of information technology
in higher education, but I make no claim to be a scholar of higher
education -- I've skimmed the journals in that area but am not deeply
immersed in them.  A journal editor might therefore send my paper to
be reviewed by someone whose anamorphic map of the literature has its
dead center in the literature on higher education, someone for whom
the literature on higher education is three-quarters of the world,
just as Manhattan is three-quarters of the world for the people Saul
Steinberg had in mind.  That limitations of that person's world view
might prevent them from fully understanding my argument, but I can
correct for that.  In fact their misunderstanding will be useful,
because it will help me to unearth the unarticulated assumption that
was leading his or her interpretation of my argument onto a different
path from the one I had in mind.  When this system is working right,
the institution can bring far more knowledge to bear on a question
than any individual could possibly bring alone.  The institutions of
the research community provide an ordered diversity, diversity within
a common framework, so that everyone gets the benefit of feedback
from people who really know the subjects that their work touches upon.

Anamorphism and feedback are also important in social and political
terms.  Two hundred years ago, people like Herder invented the idea
that people's ways of life are sorted into discrete cultures: German
culture, for example, or French, or Chinese.  The historical context
of Herder's thought makes clear why this idea made sense: Germany at
that time was politically fragmented, and the idea of a unified German
culture was part of the political movement that led to a politically
unified German nation.  Other nationalist movements found the idea of
a unified and discrete culture appealing as well, and in fact Herder's
ideas were anticipated in large part by an author in another fragmented 
not-yet-nation, Giambattista Vico, who wrote in Naples.  This idea of
discrete cultures, however, has had unfortunate consequences.  If each
culture is an organic whole that expresses its totality in every word
and artifact, then overlap does not exist.  Herder did believe that
it was possible to understand another culture, but only by getting
the entire culture into one's head through extensive scholarly study.
It must be said that there is some reason for skepticism about the
possibility of communication between cultures, given the capacity
of "civilizations" to stereotype one another to such an extent that
they don't even *care* to communicate.  But the empirical fact is
that cultures are not discrete.  German and Dutch cultures emphasize
their differences so strenuously precisely because they have so much
in common.  There really are common themes among the Meditteranean
cultures -- overlapping elements that different subsets of the
cultures share.  And the same is true for almost any geographically
adjacent cultures around the world.

Cultures, in other words, are really overlapping bundles of traits
rather than organic wholes.  Cultures do work to integrate their
various traits, but in practice they can exist in contradiction and
tension as much as in organic unity.  A culture is better understood
as a repertoire of themes, some of which are consistent with one
another and others of which are not.  A culture's repertoire is always
available to its members, who appropriate whatever themes might be
useful for them for a given purpose, and the various themes get coded
and recoded through various movements and disputes over the course
of centuries.  Once we understand all of this, the idea that cultures
are hermetically sealed from one another becomes less defensible.
Identity politics starts from this assumption of separate spheres, so
that every culture must be seen to have its own variety of science and
politics and everything else.  Fortunately, the intellectual leaders
of identity politics -- if not the routinized identity movements
themselves -- have gotten beyond this simplistic view of cultures and
identities are separate worlds.  By acknowledging both anamorphism and
overlap, it turns out, one can be oneself, value others, and presuppose
an extensive basis for communication and cooperation, without fearing
the return of a false conception of universality -- the impossible but
easily imagined idea of a perfect and complete map.

Anamorphism and overlap also help institutions to regulate themselves.
The legal system, for example, only works if every law lies at a point
of overlap of many different parties, each with different kinds of
interests.  Most especially, each law should be monitored by diverse
interests who care mainly that the law be rational -- for example that
it be applied consistently and logically, without indefensible double
standards.  Why is this abstract principle of rationality in anybody's
interest?  Because they have other laws that they care about on a more
substantive level, and they want those laws to get applied in the way
they want.  They need legal protection to do business, for example,
and they need to make sure that the legal system keeps working to that
end.  Now, of course, the legal system doesn't always work in this
way.  Every law tends create coherent classes of people whom it affects
asymmetrically, and those people will always try to pull against that
particular law on a substantive level in one direction or another.
The dangers of corruption are great, if only intellectual corruption,
and that's why it's important to have a large variety of third parties
whose interest in the issue is more abstract.  People who are affected
by the law will always form large coalitions, for example to make it
easier or harder to file class action lawsuits, but the system will
only work correctly to the extent that even more players retain an
abstract interest in the outcome of the coalitions' struggle remain
rational, regardless of how it ends up substantively.

The same principle applies to every other institution.  John Commons
points out that every institution has its own rules and its own
informal mechanisms for enforcing them, whatever formal mechanisms
it might also have.  Institutions socialize people into their values,
or at least into their language and practices.  And so long as most
participants in the institution retain a stake in its functioning,
they will act on their socialization to spontaneously enforce the
institution's rules.  Again, this is not some kind of law of nature,
and we must inquire in every case to determine whether and how well
it works.  If everyone is engaged in log-rolling then a new layer of
rules will emerge to regulate the processes by which people allow one
another to bend the official, public rules.  Institutions that depend
on representation, delegation, and agency tend to suffer from this
sort of institutionalized log-rolling, but the result may well be more
efficient, and certainly more orderly and thus predictable, than any
known alternative.  The point is that an institution's functioning is
dependent on the concerted interests of many parties whose standpoints
on a given issue differ but overlap.

The principles of anamorphism and overlap are hardly sufficient to
explain every aspect of institutions and their functioning.  But they
do suggest ways to assess institutions and perhaps to improve them.
Does the institution socialize people to cultivate an anamorphic map
of the relevant world?  What are those maps like?  How focused or
broad are they, and how does the degree of focus or breadth depend on
an individual's location?  How different are individuals' maps?  Are
they randomly or systematically different?  What incentives do people
have to map the world?  Are there points of low overlap in the world,
such as borders between nations whose citizens know little about one
another?  What kinds of mapping tools do the people have?  What roles
do informal contacts play in extending people's maps, and then what
roles do formal mechanisms like journalism play?  Are there adequate
mechanisms for drawing a diversity of people with overlapping maps
into the deliberations over a given issue?  How do people even find
out when issues arise that fall in the middle ranges of their maps --
not 9th Avenue, perhaps, but not the single cactus in the West either?
Do overlaps in people's maps serve as the basis of systematic methods
for building social networks?  If a citizen wants to know about topic
X, how easily can s/he find another citizen with an overlapping map
(so as to facilitate communication) that also includes X (so as to
facilitate learning)?  Is it worth trying to make the maps explicit?
That way people could search for one another by their pattern of
knowledge.  Do professions encourage their members to develop maps
that are too similar and not developed enough outside a parochical
boundary?  How diverse are people's maps?  Does everyone effectively
choose from a dozen stereotyped maps, or does every individual end
up with a unique map as a result of their unique interests and life
experience?  What kinds of overlap is it useful for everybody's map to
have?  Is intellectual diversity a scarce and dwindling resource, or
do modern knowledge institutions actually promote increased diversity
despite the leveling effects of global media and telecommunications?

What consequences do these phenomena have for the design of digital
libraries?  How can we conceptualize anamorphism and overlap without
falling into the twin extremes of pretending that everybody knows
everything or that everybody knows nothing?  How can people design
their own maps, aside from choosing the electives they take in school?
Can anamorphic maps be rationally designed?  Is it possible to work
backward from life and career goals to the design and maintenance of
such maps?  Is it possible to develop social networks that provide
access to people whose maps are complementary in the most useful ways?
Can the rationalization of anamorphic maps become an instrument of
social control?  Is it good enough to have diversity in a standardized
framework, as for example in the case of the research community, or is
it also important for different people's knowledge to be organized in
quite different institutional ways?

Good questions.

Okay.  Having sketched my intuition about the substantive analysis of
institutions, let's stop and appreciate my new phrase: "the principles
of anamorphism and overlap".  Doesn't that sound impressive?  A long
time ago I figured out that I could think better if I made up words
and phrases to name every intuition that started taking form in my
notebook.  The very act of putting a name on an idea causes it to
take form.  It causes me to notice examples of it, because you can
only see things that you have names for.  It encourages me to multiply
questions about it, and compare it and contrast it to other ideas,
and so on.  So I teach this to students.  In fieldwork classes for
example, I send them out to interview, encourage them to explain what
they found interesting, and then we put a name on it.  Sometimes I
compel them to make up their own name.  They find this odd, because in
their experience names are just there, the taken-for-granted gift of
authorities, and not something that anyone can make up for themselves.
You too, I say, have a right to put names on things.  In fact that's
one of the main ways that we make ourselves useful as scholars: if we
observe something and name it, then other people can observe it too.


A company called Whispercode is reported to be marketing a device that
detects an inaudible identifying sound embedded in the audio tracks
of television and radio commercials.  The idea is to measure audiences
more reliably by issuing the devices to a sample of people who promise
to keep them in their pockets.  I think this is a technology with
real promise.  Let's embed an inaudible identifying sound in the roar
given off by a leaf blower.  Then we can design a device that detects
the identifier and automatically collects a fee from the leaf blower's
owner.  This would be a free-market way of internalizing the negative
externality associated with noisy leaf-blowers, which are a pure case
of cost-shifting (from quiet, expensive brooms to noisy, cheap blowers).

We could do the same thing with car alarms, except that we'd create
an incentive to bump into a car just to collect the fee.  Still, I'd
like to see car alarms equipped with a device to report to a central
board every time they go off.  Then the local newspaper could print
a monthly top-100 of the cars whose alarms have sounded most often.
Another, lower-tech approach would be to make it illegal for one's
car alarm to go off for no reason.  Any cop who sees your car alarm
go off for no reason could just write you a ticket.  I think a lot of
tickets would get written.  I swear that I see a car alarm go off for
no reason -- not just hear it, but literally see a car that happens
to be in my visual field suddenly start screeching for no reason, or
no remotely good reason -- at least once a week.  Pure cost-shifting.


A Y2K prediction: The perfection of civil society

For another installment in my series of predictions for the year 2000,
consider the following passage from Raymond Williams, The Year 2000
(Pantheon, 1983):

  ... one of the major benefits of the new technologies could be
  a significant improvement in the practicability of every kind of
  voluntary association: the fibres of civil society as distinct from
  both the market and the state.  Today, though the dominant lines of
  communication and organisation are powerfully and centrally funded
  and controlled, millions of people, continually and irrepressibly,
  set up their own organisations, either for purposes ignored or
  neglected by the established forms, or as means of positive support
  and influence.  Typically they now work under serious difficulties,
  of resources and especially of distance.  An association can have
  a hundred thousand members and yet not more than a few hundred,
  and often only one or two, in any particular place.  The consequent
  problems of travel and funding are then devotedly addressed, but
  for many purposes the new interactive technologies could transform
  them by providing regular facilities for consultation and decision
  from people's own homes, workplaces and communities.  In any formal
  organisations, such as parties and trade unions, such facilities
  would greatly assist the improvement of democratic communication
  and decisions.  But there would also be a great strengthening of every
  kind of voluntary and informal association, from special interests
  and charities to alternative and oppositional political and cultural
  groups.  This could be, in practice, the achievement of full social
  and cultural powers by civil society, as opposed to their appropriation
  or marginalisation by the corporations and by the state (page 150).

What I find most striking about this passage is how up-to-date it
seems as a prediction.  It's still very much the way that democratically-
minded people think, for example with the emphasis on civil society,
decentralized lateral connections, the overcoming of distance, and
so on.  I've written such things myself.  But is it true?  By its
nature it's a hard thing to test: it anticipates a multitude of diverse
and mundane effects rather than anything visible and easy-to-measure.
Decisions are to be improved, associations to be transformed.  Fine, but
then there's that last bit: 

  ... the achievement of full social and cultural powers by civil
  society, as opposed to their appropriation or marginalisation by
  the corporations and by the state.

Williams was a socialist, and this (and not any sort of Soviet-style
centralization) is what socialism meant to him.  Indeed that's what
it means to most socialists.  And to the extent that he was predicting
the rise of socialism, he was of course wildly mistaken.  But then we
was not really predicting anything.  He didn't believe in predictions,
and certainly not in predicted futures that were supposed to be driven
by technology.  He held that technology is shaped by society, that no
technology is inevitable, and that people who believe that particular
technologies are inevitable are (consciously or not) covering up the
interests that such technologies would serve.  So even though he was
assessing the prospects for a happy 2000, he understood that the 2000
we'd eventually get would be a matter of choice.  Would his preferred
scenario for 2000 even be practical?  Probably not; at least it depends
what you mean by "full social and cultural powers".  The institutions
of high technology (of which the computer industry is just one corner)
are so sprawlingly complicated that it's hard to imagine them being
organized on a voluntary basis.  Yet it's striking to see the broad
and deep global consensus that a strengthened civil society is the way
of the future.  He got that right.


On the supposed democratizing effects of the Internet.

Having refuted a common argument of the Internet skeptics last time,
it is time once again to refute an argument of Internet enthusiasts.
You've heard it -- it goes like this:

  The Internet is intrinsically supportive of democracy because it
  empowers individuals, and in particular because it equalizes power
  between individuals and institutions (business, government, etc).
  Online discussion groups are a realm where the individual reigns,
  and anyone can start their own online publication on exactly the
  terms as Time Warner or the Feds.  How can you deny this when your
  own mailing list is a prime example of it?

This argument is utterly false.  To say that the Internet "empowers
individuals" by ambiguous.  On a weak reading it means that the
Internet is a tool that individuals can use.  That's true, but it's
also true of a thousand other tools.  So it's only a useful assertion
on the strong reading, where it means that it empowers individuals
relative to business, government, or whatever.  And an even stronger
version of this is what the second clause argues: that the Internet
literally equalizes power between individuals and famous institutions.

I find these latter assertions absurd.  What is the evidence for them?
It's an argument from formal equality: Time Warner can send a message
to a newsgroup and so can I; the Feds can publish their views on the
Web and so can I.  But the problem with this argument is obvious: Time
Warner has 10,000 times the audience that I do, and the Feds have the
resources to maintain 10,000 times as many Web pages as I do.  Lots
of things are like this.  I have just as much right as George Bush or
Al Gore to run for president, except that they have 10,000 times as
much money as I do.

It will be argued that the Internet does amplify the powers of the
average individual.  The Internet has indisputably made it easier to
run a newsletter with 5000 subscribers than it used to be.  But the
Internet enthusiast's argument needs something much larger: not simply
that individuals' powers have grown in absolute terms, but that they
have also grown relative to the powers of corporations, governments,
and other earthly powers.  But this is far from clear.  After all, the
Internet empowers those institutions as well: it helps them with their
communications, their bookkeeping, their logistics, their surveillance
of their employees, and so on.  It provides them with new channels for
their propaganda, helps them move quickly in emergencies, and enables
them to coordinate their people over wide areas.  So the individual
side and the institutional side are both changing, and you can't tell
who's winning until you measure them both.

I find it particularly striking that the argument I paraphrased is
often used to support two logically opposite conclusions.  Some people
argue that the individual-empowering effects of the Internet promote
democracy, but other people argue that they *suppress* democracy by
preventing governments (democratic or otherwise) from intervening in
markets.  Both of these arguments cannot be right.  In fact I think
that they are both seriously false.  (One might also argue that the
Internet will bring a new balance between democracy and markets, but
that would be a diffferent argument, if almost as implausible a one.)

So what's wrong with the Internet-brings-democracy argument?  Well,
to start with, it locates the problem of democracy in the wrong place.
Democracy requires individual rights, but it also requires people to
form larger units than themselves, including both informal networks
and formal associations.  The Internet does have democratic potential,
precisely by facilitating these sorts of connection among people.
But, and this is the second problem with the argument, the Internet
has other potentials as well.  The Internet facilitates all sorts
of societal forces, and we have no way to add up the vectors and
determine a single inevitable resultant.  And this leads us to
the third problem with the argument, that the Internet doesn't "do"
anything.  People either use the Internet to do things, or they don't,
and what people choose to do will be shaped by culture, consciousness,
historical memory, markets, and the material conditions of life.


The Internet, you may recall, is supposed to cause the "death of
distance".  Is it true?  Well, one test would be real estate prices
in San Francisco.  People in the computer industry are affluent and
good with technology; their work can be done anywhere; they should
be the very first to disperse to Caribbean islands and teleconference.
If distance is dying then we should see real estate prices in San
Francisco fall to a reasonable level as computer people scatter to
the places they'd really like to live.  On the other hand, if we see
San Francisco real estate prices go completely insane then the death
of distance is much exaggerated.  Someone might want to have a look.


Hype futures.

The word in the media right now is that business-to-business electronic
commerce, aka B2B, having once been fashionable, is o-u-t out.  This
is frustrating.  It seems to me that media discourse about emerging
information technologies has largely been captured by the small world
of short-term stock speculators -- a category that includes a lot of
venture capitalists.  The question for these people is not whether a
given category of technology is fundamentally important, or whether it
will have an important long-term impact, or even whether the technology 
works on a basic level, but simply whether stock in the current crop
of companies claiming to do business in that particular "space" can
be sold to a greater fool for a large profit two quarters from now.
These people don't deal in technology; they deal in hype futures.
So 60-day B2B hype futures are down 3/16ths on the Chicago Mercantile
Exchange this morning, and that's all they need to know.

I think it's fine, or at least inevitable, that people like this
exist.  What bothers me is when their way of thinking suffuses the
public and media discourse of an important industry.  The nature of
B2B commerce is a matter of real consequence.  All citizens should
know about the moral and legal issues that new B2B intermediaries
create, for example when they have been shaped by a small number of
large purchasers at the expense of a large number of small suppliers.
And all business people and many types of ordinary workers should
know about the structural consequences of such technologies for their
industry.  But these are not 60-day issues.  It's in the nature of B2B
commerce that it will not happen right away.  It's a long, hard slog.
It requires large organizations to fundamentally change the way they
do business.  It requires incompatible businesses practices to be made
interoperable with one another.  It requires major security issues
to be worked out.  It requires a whole layer of policies, strategies,
contracts, lawsuits, and other gnarly social stuff to be fought through.
And it's impossible for anyone to get the big picture of these major,
major issues based on the endless torrent of overhyped, distorted
"announcements" and the speculative frenzies of the stock-flippers.

We may think that the Internet stock bubble popped in April, but the
basic ways of thinking and talking that made the stock bubble possible
are still very much with us.


Research, teaching, and the professionalization of everything.

The most common complaint against the research university is that the
faculty value research over teaching.  This complaint has been current
for almost a hundred years, and nearly everyone is familiar with it.
Larry Cuban's "How Scholars Trumped Teachers: Change Without Reform
in University Curriculum, Teaching, and Research, 1890-1990" (Teachers
College Press, 1999) is a history of the medical school and history
department at Stanford that is organized around the standard complaint
and takes it for granted.  He sketches the endless cycle of attempts to
reform the university, each of which produces superficial and largely
symbolic results without ever changing the institution at a deep level.
He remarks in particular on the long-term trend toward graduate-school
research values trickling down into undergraduate classes.  I want to
dispute the premise: I think that the main problem with the research
university, and in fact with nearly all education, is that research
values do not shape education enough, and that true reform would mean
making all education much more like research.

Once upon a time, education meant cramming a lot of knowledge in your
head.  This assumption about education has two completely different
sources: the oral culture of the middle ages, for which all learning
started with memorization of canonical texts, and the industrial or
vocational model, for which all learning consists in the acquisition
of concrete, practical, modular, economically valuable job skills.
The hidden assumption in each case is that one leaves an educational
institution as a fully educated person.  This may once have been true,
and it may still be true in a way.  But many people have observed
that the world has changed.  The big change is often formulated in
terms of "lifelong learning": the notion of education does not change,
but it is asserted that knowledge keeps multiplying, and that people
therefore need to keep cramming more knowledge (aka "human capital")
into their heads for the rest of their lives.

Yet the "lifelong learning" picture is not enough either.  A better
place to start is with the professions.  Professions are not just
guilds that have quasi-official monopolies on the practice of certain
kinds of knowledge.  They are also social machines for the production
and transfer of new knowledge.  Professionals advance by pioneering
new practices, and professional conferences provide forums in which
these new practices can be presented.  A professional's public persona
consists largely of the ideas that s/he is thought to have pioneered,
and credit for new ideas is assigned both formally and informally
through peer review (and not, say, by some outside authority).  The
professions are analogous to research communities in this way, and
the analogies run deep.  Research communities have more formalized
methods of publishing their work.  But the basic method of peer review
is the same.  It's a good method, flexible and adaptable.  It creates
incentives to spread new knowledge around, and it also discourages
redundancy and encourages novelty by assigning credit to innovators.
(Being a human process, of course, it does none of this perfectly.
In fact it's a mess.  But it works extremely well on the whole.)

The Internet makes it possible to spread the professional model
to more kinds of work.  Take any occupational category at all and
imagine making it into a profession, or take any existing profession
and imagine making it more fully into a profession by supporting the
publication of ideas.  This transformation may take nonobvious forms.
Think, for example, of the management consulting firms that require
consultants to contribute case reports to a firm-wide knowledge base,
for example using Lotus Notes.  (Never mind Notes' less-than-graceful
transition to the Internet.)  From the firms' point of view, the
appeal of this practice is obvious: it makes each consultant smarter
through the aggregation of the other consultants' smarts.  The first
consulting firms that tried it, however, found that incentive systems
had to be changed.  Consultants live to be promoted, and if they can
get promoted by hoarding knowledge (thereby differentiating themselves
from their fellow consultants) then they won't want to contribute to
any knowledge bases.  Having had this pointed out to them (by Wanda
Orlikowski's celebrated paper on the subject), the consulting firms
make contributions to the knowledge base one of the criteria for
advancement.  When this is done right, consultants will strive to
formulate their contributions in the most useful way.  For all its
utility for consulting firms, this scheme differs from a profession
in that the knowledge is all proprietary.  It's not like a IEEE
conference, where the knowledge is all displayed to the public.
But it's at least one example of an occupation becoming more like a
profession through the adoption of mechanisms similar to those of the
research community.

A similar model could be applied in many other contexts.  Imagine all
of the people in the world who operate a particular kind of machine.
Then imagine a digital library in which they can accumulate what they
know.  If they don't want to write a paper then they can record a bit
of video.  (Xerox PARC has done things like this.)  Each "publication"
would be obligated to point at all of the previous publications that
cover similar topics, and new publications would only be entered
into the library if they pass peer review.  Everyone would develop a
vita that lists their publications, and formal and informal incentives
would reward long vitas.

Now, labor unions have a long tradition of resisting this sort of
thing.  Like the consultants, they have a point: if your skill is
the source of your unique value to the company, you don't want some
knowledge vacuum cleaner to come in and suck that unique skill out of
you.  After all, what's to stop them from sucking out your knowledge,
transferring it to someone making $2 a day in Mexico, and sending you
home?  The answer is that professionalization has to be more than just
alienating your knowledge.  Professionals need a degree of autonomy,
a promotion path, a culture of their own, a degree of symbolic respect
from others, and much else.  Indeed, for all these reasons corporations
regard professionals as being hard to manage.  Corporations demand
loyalty, but professionals' first loyalty is often to the profession.
After all, it's the profession that reviews their work, and it's
through professional networks that new and better jobs are likely to
be found.  On the other hand, professions help corporations by doing
a lot of quality-control that the corporation itself may be unable
to do.  That's one reason why corporations often allow their research
people to publish in the open literature.  (Another reason is that
otherwise the research people would leave.)  It's an eternal tension, 
but if professions didn't serve at least some corporate purposes
then they would have disappeared a long time ago.  They may even be
disappearing now, as more and more industries realize economies of
scale and move toward monopolies.  Medical care is an example, and
medicine is a good example of what unionists somewhat hyperbolically
call the proletarianization of the professions (see, for example,
Charles Derber, ed, Professionals as Workers: Mental Labor in Advanced
Capitalism, Boston: Hall, 1982).

The point, then, is that the Internet can be one part of a broad trend
toward the professionalization of many occupations.  Many occupations
cannot afford conferences or the cumbersome writing of research papers,
but the Internet makes these mechanisms less necessary.  Professions 
can now multiply, including many thousands of microprofessions whose
members might be dispersed around the world.  Such a development, it
seems to me, would have profound consequences for education.  School
mostly teaches people how to be in school, and so school should be
organized in ways that are analogous to the ways of the post-school
world.  In a professionalizing world, it follows that school should
be modeled on the professions.  Instead of writing term papers that
only a professor will read, students should be publishing their work
on the Internet in peer-reviewed student journals.  Professors would
probably be included among the peers, as would students in later years
of school.  Students would learn a structured framework for reviewing
their peers' work, and reviews would be graded.  Each course would
have a single conceptual framework for all of the students' projects,
just so teachers could achieve economies of scale in teaching a whole
class of students at once, but then each project would apply that
framework to a different case.

This model could be applied in a simple and gentle form as far down
as the third grade, and if it were introduced at early stages then
it could be amplified in later grades.  (Anne Gilliland-Swetland and
Yasmin Kafai in my school did something like this with kids putting
their field biology projects in archives.)  By the time college rolled
around, every student would be accustomed to having a public voice,
to having their work judged (and misjudged) by others in a public
space, to choosing topics that fit with their interests and skills,
to relating their topics from those of others before them, and so on.
These would be invaluable meta-skills that any student could then take
into any professionalized job.  Indeed, students who are accustomed
to this model would create a huge force toward the professionalization
of whatever jobs they went into.  They would take for granted that
they are creators of knowledge, not passive recipients and mechanical
applicators of knowledge.  They would take for granted that they were
part of a knowledge community, not isolated individuals who are graded
by an arbitrary authority.  And they would take for granted that they
had significant autonomy to frame their own topics in ways that served
their careers.  Of course they would also have to deal with bosses and
customers, but they would internalize the tension and balance between
getting the job done and contributing to knowledge.  The world would
be a better place.

None of this will happen simply.  Educational institutions will have
to change.  Professional communities will have to be established on
every level, and this will include creating the infrastructure -- ten
thousand digital libraries, each geared to the needs of a particular
subject matter and level of learning.  But because the institutional
mechanisms within all of these communities will be so similar, it
stands to reason that the task will get much easier once the first
hundred communities have gotten established.  For the educational
institutions, the biggest challenge will be managing these new matrix
connections.  No longer will each class proceed in relative isolation
from counterpart classes in other schools.  Instead, instructors and
schools will form alliances according to their way of teaching the
material.  More quantitatively oriented schools might form one online
professional community, where more qualitatively oriented schools
might form another.  This is how it works with the research world,
which is self-organizing largely because anybody is welcome to start
a journal if they can find enough people to write for it and read it.
Starting a journal is such a good career move that (in my experience)
new journals are started whenever they can be.

Likewise, starting a new professional community of peer-reviewed
publication will be a good professional move for a teacher in the new
order.  Teachers will finally have a career path of their own, as they
apply this same model to their own work, as well as supporting their
professionalized students.  Teaching will become more like research,
and as peer review of teaching activities becomes widespread, teaching
will start being valued as much as research.  The standard complaint
against the research university will have been addressed, but in
precisely the opposite way from what's usually envisioned.


In case we haven't beaten to death the question of Steve Ballmer's
audacious claim that nobody has ever called Microsoft untrustworthy,
one loyal reader faxed me a couple of pages from Stephen Segaller's
"Nerds 2.0.1: A Brief History of the Internet" (TV Books, 1998).
Recounting a disastrous joint venture that 3Com entered into with
Microsoft under Bob Metcalfe, Segaller says (among other things):

  Metcalfe resigned from 3Com, after eleven years in which the
  company had grown from one employee to two thousand, from zero to
  $400 million a year in sales.  As it turned out, that was just the
  rehearsal: the growth show was yet to come.  But it left Metcalfe
  with a deep dissatisfaction with how Microsoft does business: "When
  I complained to Microsoft, the guy involved, whom I will not name,
  said 'Your mistake was, you trusted us'."  (page 255)

Bob, as you probably know, is not exactly an antitrust hawk.


Some URL's.

UXN Spam Combat

Bombs Missed Kosovo Targets

Making Sense of the Wireless Internet,194,5,00.html

SIAA brief urging the Supreme Court to hear the Microsoft appeal

Human ID Chip Implant Prototype Unveiling

Agency Could Be Coming for Your Domain Name,176,474,00.html Drops Tracking Service Amid Pressure

police violence in Philadelphia

full text of judge's ruling against Napster

British Library Junks 80,000 Books,3604,353004,00.html

News and Information for Monster-Hedge Victims in the UK

Design Education 2000 Conference

Truth Be Told

Richard Ling's ethnographic studies of cell phone use

Relate-Create-Donate: An Educational Strategy