Some follow-up on recent topics and notes on (de)centralization and
professionally twisted language, plus an excellent batch of URL's.

I am still far behind on my electronic correspondence, and I apologize
to all the people that I have not gotten back to.

How do I find time to write this stuff?  Easy.  I write it when I am
supposed to be writing something else and can't.  It clears my head
out and keeps the words moving.  It's all first draft, with little
editing, which is why I keep having to go back and clarify bits and
pieces in response to subscribers' correspondence.  But at least it
gets written.  I think the world would be better off if I could make
my living just doing this.  It would certainly be more efficient.

I'll start by cleaning up several points that people have raised
in response to things I have written in the last couple of months.
Also, several people have sent me URL's for Web pages that pertain to
recent topics, such as the massacre victims now being dug up in Kosovo
and the method by which companies in the computer industry should
account for employee stock options.  Even though some of these Web
pages qualify or disagree with my own suggestions, to make my life
easier I have simply listed them at the bottom of this message without

In my little essay on the WTO protests in Seattle, I said a couple of
things that weren't real precise.  One of them was:

  Measures that supposedly produce a decentralized society will in
  fact reliably and predictably accomplish the reverse.  Am I saying
  that we should shut down capitalism and return to the stone age?
  No.  Markets work when they work.  We should just stop lying about
  what markets are and what globalization is.

I felt bad about the word "lying".  It's a word that I use when it is
appropriate, but it was not appropriate here.  In fact it's an example
of the kind of language-twisting that I am confronting elsewhere: it
inflates the word to cover things that it hasn't and shouldn't cover,
in this case (a) matters of judgement and theory, and (b) matters that
the alleged liar presumably believes to be true.  In contrast to the
speakers of the fashionable jargon who really do seem to be acting in
systematic bad faith, I believe that most libertarians believe what
they are saying, even when the things they say seem (to me) obviously
to be the opposite of the truth.  In the same essay I also said this:

  ... Global networks are not a force for decentralization.  To
  the contrary, our lives are increasingly mediated, structured,
  monitored, and regulated by electronic systems that are controlled
  by highly centralized firms. ...

Some took this to mean that global networks never participate in any
kind of decentralization.  No, I just meant that decentralization
is not a reasonable description of the major trend in the world that
global networks participate in, and that they participate in some
gigantic trends to the contrary.  There are good reasons for this that
I have explained on many occasions, but which somehow never become

In that same essay, I referred to the nutcase vandals in Seattle as
anarchists, which made real anarchists unhappy.  I was wrong.  Not
that I really approve of anarchism as a philosophy, but the people
breaking the windows at Niketown were cartoon anarchists, not the
genuine item.  (By the way, eyewitnesses tell me that the looting
they saw was caused not by political people but by just plain looters
who took the opportunity opened for them by the police riot and the
cartoonistas.)  I also referred to the nutcase vandals as violent,
but some people feel that the word violence should not be applied
to attacks on property, only attacks on people.  I am sympathetic to
the argument that a broader use of the term tends to equate morally
dissimilar situations, but I am less sympathetic to any suggestion
that trashing Niketown is a legitimate form of civil disobedience.

Finally, people keep asking me what problem I have with paragraph
breaks.  I don't know, and I don't spend much time wondering about
it.  It's just one of those things that I don't feel like doing, and
the only reason this list works is that I only do what I feel like.
You're welcome to insert paragraph breaks when you forward a message.

When I was writing about the institutional dynamics of out-of-control
academic jargon, I said the following about the work of someone who
was trying to be Gilles Deleuze:

  It's all very portentous in such a way that graduate students often
  feel obliged to pretend that they understand it.

A couple people got offended by this.  It wasn't real clear.  First
of all, no, I did not have anyone in particular in mind.  If you
thought I had you in mind then I'm certainly sorry.  More generally,
my point was not to abuse graduate students, who are among the world's
finest people.  Let me explain the point a bit less telegraphically,
expanding on my earlier argument, because I think it is interesting.

New graduate students face a whole series of institutional problems
that are hard to explain to someone who hasn't already mastered the
workings of the institution.  One problem is endemic to human life
in general, namely that you're always entering conversations in the
middle.  You show up someplace -- a new job, perhaps -- and the people
there already have a conversation going on.  They probably have quite
a few running conversations, and they have probably accumulated a big
network of shared background assumptions.  Many words have probably
acquired specialized local meanings whether the people are aware of it
or not.  Meanings will have been shaped by long-past events (what the
anthropologists call "critical incidents") and political fault-lines
that nobody ever needs to mention.  Even an innocent word choice can
place you on one side of a conflict or another.  These phenomena need
not be spectacular or pathological, but they are certainly universal,
and they can seriously confuse a newcomer.

One way to understand academic language is that this entering-a-
conversation-in-the-middle effect is amplified about twenty times
relative to any normal setting.  That's because academics are paid
to say things that are new, which is very hard, so that they are
continually torquing their language -- usually for good reasons,
but of course not always.  As a result, new graduate students can be
forgiven if they feel like they are walking around in a linguistic
minefield.  What's more, the language that graduate students encounter
in professional settings is a kind of capital.  That is, the ability
to use the language is a valuable commodity.  Talking a specialized
academic language is what one gets paid to do, or at least it's a
precondition of what one gets paid to do, which is hopefully to say
something, and so it is understandable if students feel obligated to
learn the languages they hear.

As a teacher, I find these things frustrating.  My first question is
always: what do you care about?  Having gotten an answer to that basic
question, we can go looking for suitable conversations to join.  But
graduate students are not stupid, to the contrary, and if Foucauldian
vocabulary is valuable capital then they can spot that fact a mile
away.  They know that the job market sucks, and they intend to get the
capital they will need to get a job.  I don't mean to overgeneralize.
Everyone is different.  Still, I often find myself saying, no, you
really don't have to learn to talk that way unless you intend to join
a conversation in which everyone else talks that way.  But that's not
how it seems when you're new and you have to graduate in five years
and you don't yet have a differentiated sense of the terrain.  Who's
really right?  Maybe they are.

Now think about the collision between these phenomena and the whole
problem (which I discussed before) of importing French philosophy to
the United States.  The French think very highly of philosophy, and
they have an exceedingly centralized and hierarchical meritocratic
system for identifying and training the best philosophical talent.
Even though they take their philosophical training system for granted
and even harp on its many defects, it nonetheless works very well,
at least relative to all of the other systems.  What this means is
that French philosophers assume an audience that is widely read and
deeply sophisticated.  That is one reason why people like Foucault
and Derrida employ very little of the conventional scholarly apparatus
of citations and surveys of related literature: they can assume that
their readers will know and recognize all of the precurors of their
ideas.  They sometimes write about one another (see, for example,
Michel de Certeau's powerful chapter on Foucault in Heterologies:
Discourse on the Other, University of Minnesota Press, 1986), but they
don't try to footnote every little idea.

This system may sound bad to American ears, but it works: it enables
these authors to get a great deal of intellectual leverage from the
background of knowledge that they share with their readers.  It is
the kind of pressure-cooker that, as Randall Collins suggests in his
stupendous book The Sociology of Philosophies: A Global Theory of
Intellectual Change (Harvard University Press, 1998), is required
for any great philosophy to get done.  It is the ongoing-conversation
effect multiplied by fifty instead of twenty, and its decline is
probably why (so far as anyone can tell) no great philosophy is being
written right now.  To get an idea of what I mean, have a look at Mark
C. Taylor, ed, Deconstruction in Context: Literature and Philosophy,
University of Chicago Press, 1986.  It is a scholarly sourcebook of
the precursors of Derrida's method of deconstruction, and it is quite
a revelation.  Derrida suddenly seems not like a Martian but like
an incremental advance beyond a whole series of people like Levinas,
Bataille, and Blanchot.  Now, serious specialized scholars in the
United States certainly understand this.  But it takes real work to
become that serious, and most people, not having been brought up in
the French system, will never have the time.

Now pick up these texts and move them to another country, such as the
United States.  I referred to the tendency of certain scholars to copy
the style of someone like Foucault in a superficial way, but I didn't
fully explain why this causes so much trouble.  Academic discourse
only works if it's part of a dialogue.  In France, philosophical
dialogue works because everyone knows the background.  Individual
authors can develop highly personal writing styles without disrupting
the conversation.  Some of these writing styles have more of a point
than others, and I choose Foucault as my example because his own
style (prior to the relatively plain language of his last few books)
was much less motivated than that of the others.  Derrida's style,
by contrast, is perfectly well motivated by his philosophical project.
So is Bourdieu's.  Lacan's style, probably the hardest of any writer
who ever lived on this planet, is also motivated by his conception
of psychoanalysis, and really only made sense in his public lectures.
Deleuze's style is motivated by a project that I think is pretty
misguided, even if many of the individual points are useful.  And from
what I can tell de Certeau's style derived more from his astounding
depth of learning than from a specific project.  (I realize that these
people aren't all philosophers.)  When Americans copy these styles,
disaster often results because the conversation is broken.  Readers
in the American context generally cannot see the language as part of a
densely organized dialogue, so the whole thing locks up.  The dialogue
loses its dynamic, forward-moving quality, and everyone falls into
a kind of intellectual autism, a black hole from which nothing can

This is not to say that Foucault, for example, has had no beneficial
impact on American scholarship.  Scholars who employ the ideas without
copying the style often have useful things to say.  An example would
be John and Jean Comaroff's multiple-volume anthropological history
of the Tswana in northern South Africa, Of Revelation and Revolution
(University of Chicago Press, 1991 and 1997).  Their research is
influenced by Foucault, but you wouldn't know it to read their prose,
which is somewhat mannered to be sure but for their own reasons and
not because they are copying anybody.  Instead of falling into a
solipsistic vortex of writing style, they have engaged with the ideas
and digested them into their own thinking, along with everything else
that they have engaged with, which is a lot.

There is one final reason why people in academia, including graduate
students, often feel compelled to acquire specialized languages that
are not necessarily suited to their own projects: academic languages
exhibit network effects.  Just as people around the world invest in
learning and speaking English because so many other people already
speak English, likewise the theoretical vocabulary of a particular
author can become the de facto standard of conversation in a certain
field.  And in case you think this is just an artefact of the fashion-
ridden humanities, you should know that mathematics is one of the
fields where it happens most furiously.  A mathematician who invents
a new formalism (what they call "machinery") will be forgotten unless
other mathematicians use that formalism to prove theorems of their
own.  Often a variety of formalisms are available that do generally
the same kind of work, one way or another.  Each mathematician has
an incentive (not necessarily overriding, especially when the choice
of machinery makes a major difference in the results one can obtain,
but still significant) to use the same machinery that everyone else
is using, precisely for purposes of compatibility.  In this way the
development of mathematics is path-dependent, with some well-promoted
or centrally-networked authors defining the basis of subsequent
development in their fields, while other authors retire in obscurity.
I don't mean to disparage the mathematicians' culture, which is
perfectly nice.  It's not about anyone's human qualities.  Network
effects happen whether people are elbowing one another or not.

The same thing is true in many other fields.  Once Foucault becomes
the vocabulary of choice for talking about the social construction of
the body, for example, people will use Foucault-speak for that purpose
even though some other author's vocabulary might be better-suited to
a particular purpose.  And just as newcomers to a field of mathematics
frequently sledgehammer a problem with machinery that is too general
to reveal its inner logic, likewise newcomers to social theory will
use five-star Foucauldian jargon to say things that could be said
using the admirably plain language of John Commons or Anselm Strauss.
Outsiders will mistake this for academic empty-headedness, and that's
sometimes what it is.  But at least as often it's more complicated.
And the humanities and social sciences get a disproportionately bad
reputation for doing it because outsiders haven't the slightest clue
what the mathematicians are saying, whereas they think they have a
clue what the others are saying.

So that's what happens.  The graduate students walk into the middle
of this very complicated set of dynamics that nobody ever explains.
That leaves the students feeling compelled to master arbitrary codes
that their careers seem to depend on.  It's that structural situation
that I am interested in, not the properties of graduate students
themselves.  Graduate students face a serious structural problem in
roughly the middle third of their careers, after they have finished
their structured coursework and before they have solidified a thesis
topic and begun presenting papers at conferences.  Until you've got
a thesis topic, the world can seem completely chaotic.  Everything is
connected to everything else, and one's candidate topics are grossly
too ambitious.  Everything is slippery and formless, and you usually
don't have enough of a social support system to provide a sense of
proportion for the process.  Many students never make it through to
the other side.

The problem here is very clear, even though it's very hard to explain
to the people who are going through it.  Graduate school is what
anthropologists call a liminal phase: a chaotic transitional period
between giving up one social status and acquiring another.  Many
people wonder why we make each student write a huge dissertation that
will sit on a library shelf unread by all but a few people.  Of course
many dissertations (not enough of them, but many) will be broken down
into journal articles or revised into books.  But that's not the real
issue.  The fact is, when you are producing your dissertation, the
most important thing that you are producing is yourself -- yourself
as a new member of a profession.  You are figuring out what you care
about, you are rehearsing a professional voice that is in dialogue
with lots of other professional voices, you are inserting yourself
into a social network by establishing professional relationships with
all of the other people whose work is related to your own, and you are
learning to fulfill your responsibility to give proper credit to all
of the people whose work anticipated your own.  Having accomplished
all of this, you might as well throw the dissertation away, because
the new social embedding that you have established is probably much
more valuable than the stuff you wrote in your dissertation, which
will inevitably turn out with time and distance to be both much
better work than you think it is while you are doing it and a ragged
first draft of a random first installment on what you really wanted
to say.  Once you have fully joined the research community that shares
your interests, you will have all sorts of feedback and all sorts of
reality-checking and collaboration and social support.  Life probably
won't be much easier, but it will certainly feel less chaotic.

The real purpose of a dissertation, then, is to help solve a chicken-
and-egg problem: you need a professional community in order to do
research, but you can't join a professional community without doing
research.  Because they cannot yet see the phenomena of professional
community, students very often, and very understandably, fasten onto
formal aspects of the process: politicking their thesis committee,
passing exams, mastering jargons, and so on.  You can't ignore that
stuff, but you can get it into proportion by focusing your attention
on the communities you think you may want to join: identifying
them, scoping them out, watching them in action, meeting the people,
figuring out whether their values match your own, and so on.  In
the old days of patriarchal academia, a student's advisor would get
him a job through his own network of cronies.  Today, however, part
of a student's job, hopefully with the help of advisors and friends
and everyone else, is to build a community around his or her work.
Needlessly esoteric academic languages give students the wrong idea:
they portray research as a matter of becoming someone else, rather
than becoming a professional version of yourself.  They make it seem
like becoming a research means acquiring someone else's voice, rather
than developing one's own.  And they exaggerate the degree to which
success in research depends on making oneself accountable to other
people's agendas, rather than actively seeking out a community of
interlocutors whose agendas can be brought into productive dialogue
with one's own.

The institutions of research are hardly perfect.  But I think that
their imperfections would be best alleviated not by blowing them
up and placing them under the power of some extraneous authority,
but rather by systematically teaching graduate students the things
that I am saying here.  Owning and applying a powerful model of the
institutional dynamics around one is where sanity begins, and it is
the best way to dissuade people from the misguided strategies that
reproduce institutional pathologies rather than dissolving them.

In response to my latest complaint about the >'s that deface messages
that are forwarded by certain mail-reading programs, one loyal reader
broke to me the sad news that this convention originates with Emacs.
Yet somehow it never seemed so bad back in the old days.  Forwarded
messages did not have to contend with automatic MIME and HTML markup,
or with automatic wrap-around of lines that were longer than so-and-so
many columns.

In response to the same piece, a long-time Internet activist suggested
with some vehemence that the decline of heavily forwarded political
action alerts has been due to campaigns such as mine against spam.
Hmm.  Many people do mistakenly equate action alerts with spam, but
I have to ask, how is that my fault?  One can spam an action alert
the same way one can spam anything else, but most action alerts are
forwarded by people to other people they know, or to discussion groups
that they regularly participate in.  And my guide to writing action
alerts warns against spam.  So at worst it's not the campaign against
spam but the misconceptions about spam that are the problem, or else
antagonistically inclined people deliberately pretending that action
alerts are spam even when they are not.  It's a cultural issue, and in
my own very small way here I try to intervene in the culture.

A couple of people observed that they still receive the old classic
hoaxes (the $250 cookie recipe, the kid who wants to receive post
cards in the hospital, the modem tax) ever time a new round of users
joins the network, and I don't doubt that those messages will continue
to behave like latent viruses in the Internet nervous system until the
last human being is completely jacked in.  I'm only remarking on the
decline of messages like that traveling very fast from one person to
another by forwarding.

One reader pointed out a counterexample to my claim about the demise
of action alerts: the successful campaign against "Know Your Customer"
banking regulations, which occurred early this year if I remember

Finally, several people remarked on my comments about getting old.
Yes, I was joking.  There's nothing wrong with being 60 or 39 or any
other age.  I don't have a huge problem with getting older.  I know
more and I'm a better person, relatively speaking.  The only problem
I have is the ticking of the clock.  And the clock is ticking pretty
fast right now.

My scenario for the end of Moore's law was wildly unpopular.  I should
make clear that, like most of my scenarios, it was just a scenario,
meant to provoke analysis rather than to stand as a prediction.  The
point was: if you believe in Moore's law, you believe in a complicated
economic and social phenomenon, not just in technical progress.  Some
people disputed whether the fixed costs of making chips are rising
quite as fast as I said, and others disputed whether those fixed costs
would remain high with new technologies.  They're probably both right.
In particular I got reports on a lot of exotic new circuit-making
technologies.  My general rule about those exotic technologies is that
I'm not impressed by a single gate, but I'm definitely impressed by
a four-bit counter.  Few of the exotic technologies are to that point
yet, but I as much as anyone think it would be cool if they got there.

The most common complaint was that I lacked imagination about what an
exponentially larger number of exponentially more powerful circuits
might be good for.  Many of the proposals involved things that people
can watch, but those proposals are limited by the number of eyeballs
in the world.  It is thus our duty to think up as many applications
of high-end computing as possible, just so we can soak up all of the
high-end chips that will be rolling off the fab lines of the future.
Unpatriotic 180MHz laptops will shunned; conspicuous computation will
be in.  Here's my proposal.  It involves an excellent cycle-soaker:
speech recognition.  We need a device the size of a pager with a
microphone and a speaker.  You attach it to a wall, or underneath
a meeting-room table, and it sits quietly waiting for someone to
utter to phrase "mission critical", whereupon it makes a loud noise
like Daffy Duck.  A fancier one could have a simple packet wireless
transponder.  You could pay $1 a year to have the Dilbert corporate
buzzword of the week automatically downloaded into it.

In my Kosovo message I inadvertently referred to Congo by its Mobutu-
era name, Zaire.  This was not a political statement, even though the
new boss is pretty much the same as the old boss.  It's sad.  At least
the people have great music to listen to.

One of my fragments on Kosovo omitted a word.  I wrote:

  Thirty years ago [Europe and North America] turned away when their
  good friends in Indonesia slaughtered people on a scale matched only
  by the Khmer Rouge.

I had intended to say, "since matched".  Hitler, Stalin, and Mao each
killed more people than our esteemed statesman-ally Suharto, even per
capita, although he was right up there.

I mentioned that some of the commenters on my "self-limiting Internet"
paper were fixated on themes of centralization and decentralization.
I had expected that these experts on Internet architecture would
set me straight on important technical details, and they did do
some of that.  But they spent most of their effort criticizing me
for advocating centralized government control of the Internet, even
though I had done no such thing.  At length it dawned on me that
their arguments were all precisely backwards.  What they called "the
market", namely the system by which the Internet arose, was in fact
the government, indeed a centralized and hierarchical part of the
government called ARPA.  And the scenario of mine that they called
"legislation" and "a central authority to direct the 'proper' outcome"
was in fact a mechanism of the market.  These are very smart people,
and yet they were contorting themselves to fit my argument into a
conceptual frame that is the opposite of what I said, and the opposite
of the truth.

I'm sure they were doing this in good faith; after all, they hardly
invented the ideas they were pressing on me.  It's just that I had
never seen so much intelligence put in the service of making those
backwards arguments.  Literary critics talk about this: it's the
really smart thinkers who bring the contradictions of a discourse
to the surface, even when -- especially when -- they don't understand
them.  The idea that the market solves coordination problems (among
other ways) by centralizing things through property rights is totally
unintuitive within the cyberspace worldview that identifies the market
with decentralization and self-organization.  But it is common sense
to economists.  And it happens all the time.  Am I advocating this
kind of centralization?  In fact I was warning against it.  But at
a more basic level, I want to get us beyond the simplistic opposition
between decentralized markets and centralized governments, and beyond
the widespread habit of berating anybody who even faintly appears
to countenance anything but an idealized notion of the former.  And
by all means let's stop tolerating the practice of calling people
"statists" when they advocate any exercise of democracy.  It's just
a way to call people communists without taking responsibility for it.

I had all this in mind as I read Janet Abbate's book, Inventing the
Internet (MIT Press, 1999).  It's a good book.  It's concise, it has
the facts, and each chapter has a nice clear point stated in plain
language.  But it too struggles with the themes of centralization and
decentralization.  You've often heard it said that the Internet arose
within and vindicates a decentralized, bottom-up, self-organizing kind
of social process.  But this story bears only a glancing relationship
to reality.  Abbate has the facts straight, but somehow loses them.
So, for example, she recounts how ARPA compelled its contractors to
use ARPANET, in many cases over their strenuous objections (pages 46,
50, 55), how it later drove the development and adoption of electronic
mail by methods such as being accessible to their contractors only
through that medium (pages 107-110), how the operational branches of
the military pushed for widespread adoption of TCP (page 133), how
ARPA funded implementation of TCP/IP on many vendors' machines (page
143), and how ARPA imposed an absolute death march on its contractors
for the transition to TCP (pages 140-142).

All of this is recounted in the most graphic terms.  Yet by the end of
the book she is referring to "[t]he Internet community's decentralized
authority" (page 182) and to its "unique tradition of decentralized,
user-directed development" (page 218).  This is odd: if anything the
Internet's development would seem to vindicate benevolent despotism.
The ARPA guys were outstanding managers; they managed not principally
by ordering people around (though they certainly did that too) but
by manipulating the social networks and infrastructure (pages 55, 69,
72-73).  They built upon and amplified the researchers' own informal,
meritocratic, and community-minded culture rather than trying to
replace it (e.g., pages 54-55, 84).  They stayed light on their feet
by spinning off every task that a normal bureaucratic agency would be
capable of performing.  And partly for this reason they were the only
game in town.  They had massive power over their contractors, and they
had no problem using it (page 55).  Yet their technical values were
so deeply aligned with those of the researchers, and their management
methods were evidently so effective, that their vassals felt like they
were in a decentralized and antibureaucratic heaven, instead of living
on ARPA's plantation.

So was there no decentralization?  This is the hard point.  Time and
again throughout modern history, the institutional structures that
facilitate decentralized social processes have been established through
the exercise of centralized power.  The metric system is an example.
Once established, it reduced transaction costs and facilitated trade,
but it could not be established without the massive centralized effort
of the Revolutionary government in France (see Ken Adler, A revolution
to measure: The political economy of the metric system in France,
in M. Norton Wise, ed, The Values of Precision, Princeton University
Press, 1995).  The market institutions of capitalism, likewise, arose
not through spontaneous exchanges of someone's barley for someone
else's cow but through the overwhelming and totally conscious design
of the state (see, for example, Karl Polanyi, The Great Transformation,
Beacon Press, 1944; or the works of Geoffrey Hodgson).  In each case,
just as with the ARPANET and Internet, the government established
compatibility within a sufficiently large domain that network effects
could take over and spread the compatibility much further.  Abbate
stresses over and over the many roles of users in shaping the ARPANET
and Internet (pages 4-5, 34-35, 64-65, 85, 90, 93-94, 111).  But she
equivocates.  She easily demonstrates that the powers that be in ARPA
were wise enough to listen to and learn from the feedback they were
getting from their contractors, even when that feedback caused them to
change drastically their understanding of the technology (e.g., page
111).  But by the end of the book, she is referring to "participatory
design" (page 220) and using other formulas that suggest that the
users had formal power over the design process, as opposed simply to
advisory input into it.

Part of the problem, of course, is chronological.  Abbate's concluding
formulations were actually somewhat true by the time her story ended.
Once the distinction between TCP and IP was discovered and implemented,
once both standards had been imposed and had achieved critical mass
-- once, that is, the technical conditions had been established for
decentralized development -- then it was possible for ARPA to spin
the Internet off, and then for NSF to hand it to the private sector.
Now the hard question -- the question that I was asking in my paper --
is whether the Internet architecture can improve qualitatively in its
current institutional organization, or whether some *private-sector*
centralization will arise to organize the same kind of transitions
that a centralized ARPA imposed on its contractors way back when.
I have no idea of the answer, and I am not advocating one, centralized
or otherwise.  I'm just doing social theory, and social theory is
about questions.  We are only going to have good useful questions if
we step out of the simplistic oppositions that shape the debate now.

What could we do with better concepts?  Well, we could productively
compare and contrast the type of centralization represented by ARPA
in the development of the Internet with the type of centralization
represented by Linus Torvalds in the development of Linux.  We could
notice that participation in both systems is voluntary: nobody ever
forced anyone to be an ARPA contractor, although once you've made an
investment in getting ARPA's money ARPA has plenty of control over
you, and nobody forces anyone to contribute code to Linux, although
once you've made an investment in building your system on top of
Linux you acquire a large stake in its future stability.  We could
also notice the role of a strong shared ideology in coordinating
the activity in both cases, the ways in which that ideology became
inscribed in the architecture of the system itself, and the subsequent
role of the architecture in the dynamics of the social system.  We
could analyze the role played in both cases by charismatic leaders,
and we could contrast each case to competitors who lacked one or
more of these conditions.  In short, we could start making some
fine distinctions among various kinds of relationships that centers
can have to the things they organize.  All of this and more becomes
conceptually possible once we start dividing the world along simple
centralized-versus-decentralized lines.

Professionally twisted language has become infinitely more widespread
in the last several years, and so here is another installment in my
project of confronting it.  Let us consider the following passage from
the cover story about Richard Nixon in the 9/20/99 issue of US News
and World Report (page 27):

  "[In response to Daniel Ellsberg's release of the Pentagon Papers,
  Richard Nixon] caused junior staffers to burglarize the safe of
  Ellberg's psychiatrist.  The daffiness of this enterprise was a
  sign that Nixon had become unhinged by the virulent opposition of
  elite liberals who were labeling as a war crime his de-escalation
  -- U.S. troop strength was falling by then -- of a war their
  party's president had, with little or no protest from most of them,
  constantly escalated."

What's most striking about this passage, as indeed about the entire
article it comes from, is just how dense the sophistry is.  It's not a
matter of casual laziness.  You can't twist language this hard without
really trying.  Rather than spending a week taking the passage apart,
though, let's just consider a handful of points.

(1) The party of personal responsibility blaming anti-war protesters
for Richard Nixon's criminal behavior.  This is a common pattern: when
men leave their families, for example, they blame feminists.  Personal
responsibility, it would seem, is for other people.  Notice too the
insanity defense, which is unfashionable for ordinary criminals but
evidently appropriate for certain of the more privileged ones.

(2) The falsehood that opponents of the invasion of Vietnam regarded
the so-called de-escalation as a war crime, as opposed to the violent
acts that still went on in very high numbers afterwards.  It wasn't
killing them less that was said to constitute the alleged war crime,
but killing them at all.  Philosophers call this "substitution into
an opaque context": putting one's own favored characterization of a
situation into a grammatical context where someone else's intentions
belong.  In this case the context is in "liberals who were labeling
as a war crime X", where X should properly be a fair paraphase of
the liberals' way of defining the thing they objected to.  This too
is a common pattern.  Another version is the phrase "Let me get this
straight", which in the current jargon is always followed by a grossly
tendentious paraphrase that pretends to summarize the other party's
thinking while recasting it in a form that the other party would never

(3) The innuendo that the antiwar protesters were Democrats who were
biased unfairly in favor of "their party's president", namely Lyndon
Johnson.  This is breathtaking, given that the antiwar protesters had
driven Johnson from office, and given that "their party" was perfectly
willing to engage in extreme violence against "them" in the streets of
Chicago in 1968.  So the way the passage works is fancy; it hinges on
the phrase "with little or no protest from most of them".  The antiwar
movement was, of course, large and vocal during the Johnson period.
As the horror accumulated and more secret information about the war
(such as the Pentagon Papers) was revealed, the movement became much
larger.  It follows statistically that "most of them" were not part of
the movement earlier, and the motives ascribed in a sideways fashion
to "most of them" can then be smeared around by association to all
of them.  What's frustrating for a rational person reading the passage
is that the author has issued a sideways insinuation about the motives
of the majority of the antiwar protesters, one that inherently cannot
be refuted, and any attempt to evaluate it or explain what is unfair
about it requires lots more words than anybody has time for.  It's
framed as an insinuation for a good rhetorical reason: if the author
came out and said, "they did not protest earlier because they were
protecting Lyndon Johnson", then a reasonable reader would respond,
"what's your proof?" and "are you a mind-reader?" and "how do you know
it wasn't for another reason, such as the newly released information?".
But if it's just an insinuation then the reader first has to establish
that the claim had been made, and only then demand rational arguments
for it.  This sort of insinuation is also common in the current jargon.
It is probably its dominant tone.

(4) All sorts of word choice.  First, "daffiness": notice how the
word tends to trivialize criticism of Nixon's actions as well as the
actions themselves; they are merely "daffy", not (say) "evil".  Next,
the biological metaphor "virulent" as opposed to some human attribute
such as "principled".  How do you refute the claim that opposition
was "virulent"?  Polemics work better if they are just vague enough
that they cannot be refuted.  We've gotten far too accustomed to
that sort of vagueness.  Then we have the phrase "elite liberals".
This is associationism.  Liberals of all sorts, civil rights leaders
for example, were strongly critical of the bombing campaign, but it
suits the argument only to pick out the "elite" liberals.  The effect
is to make it sound, without ever having to say so, like all liberals
are "elites".  At an even more basic level, sneering at "elites" is
dishonest unless one is willing to sneer at bankers and senators and
people who write cover stories for US News and World Report as well.
Finally, the word "labeling", which carries no more meaning than a
word like "calling" except to suggest in a vague way that something
was inherently wrong with it.  It takes much more effort to explain
what's arbitrary about this sort of word than it does to use the
word.  The word does have legitimate uses, of course, but only when
accompanied by some clear account of what the problem is.  It's the
same effect as with the technique of insinuation: the ratio of words
required to unpack it to the words required to do it is large, and
the sheer accumulation of such high-ratio rhetorical tricks in the
passage means that the audience is long gone before anyone could
possibly unpack all of the sophistry it contains.

Let's try another one.  This passage from the Wall Street Journal was
quoted in the 11/29/99 issue of Forbes (page 44):

  If Microsoft cannot act rationally in its own interest, the
  alternative is a government administrator to take over the business
  and run it for the benefit of Microsoft's competitors.  Outside a
  Nader thought-bubble, there can't be many people who don't see this
  cure as worse than the disease.

The first sentence here is plenty.  Notice the rhetorical devices by
which the precisely drawn issues of the Microsoft case are inflated
into the broadest possible exaggeration.  There is a very legitimate
conversation to be had about exactly which kinds of conduct are anti-
competitive in software markets.  But the WSJ chooses to draw the
issues much more widely: whether Microsoft can "act rationally in its
own interest".  It can be stipulated that the actions by Microsoft
that are at stake in the trial were rational, and that they were
in Microsoft's interest.  But beyond that point the WSJ's grammar is
ambiguous, following the some/all pattern that I described last time:
does the trial concern only the rational actions that the government
had enumerated, or does it concern all possible rational actions?
Well, clearly the suggestion is the latter: the government is said
to propose that Microsoft not be allowed to act rationally in its
own interest at all.  Otherwise the rest of the sentence does not
follow.  But it is obviously false that the government is proposing
such a thing, so the other interpretation remains open as a fallback

On the flip side, the WSJ would seem to suggest that Microsoft, and
by extension any business, should be able to take any action that is
rational and in its own interest.  But this is plainly false as well;
after all, the small businessman who stuck a shotgun in my face one
evening and instructed me to hand him my wallet was acting rationally
in his own interest: he got away with it.  So the WSJ is now on record
as supporting anarchy, at least so far as businesses are concerned.
Ralph Nader and the government are then put at the opposite extreme.
But are Nader and the government really arguing for "a government
administrator to take over the business and run it for the benefit
of Microsoft's competitors"?  Of course they aren't.  Even the most
extreme of the proposed remedies, breaking the company into parts,
whatever one thinks of it as policy, does not fit that characterization.
A slightly less extreme interpretation is that the WSJ views any
constraint on self-interested rational action as being tantamount to
a complete government takeover.  But that is not true either, as would
become clear if the the gentleman with the shotgun were to choose the
WSJ editors as his next customers.  It's a ridiculous exaggeration.

So what's going on?  The passage depends for its rhetorical force on
its ambiguity.  Without the ambiguity, the sentence would be obviously
false no matter what.  But with the ambiguity, the mind gets lost
trying to follow all of the logical pathways, and somehow the sentence
ends up making a dim kind of sense.  What kind of sense?  The kind
of sense that you understand with your lizard brain: broad, vague,
emotionally primitive dichotomies.  Do the editors of the WSJ operate
solely from their lizard brains?  It's an empirical question.  Either
they understand what they are doing, or they don't.  You decide which
is worse.

Staying with the Wall Street Journal for a moment, let us consider
Peter Huber's 12/20/99 editorial page article about the class-action
antitrust lawsuit that agricultural activists filed last week against
Monsanto.  The issue, briefly, is that Monsanto has patented a couple
of genes and, it is argued, is trying to establish market power by
engaging in anticompetitive licensing practices -- I assume, though
I haven't read it, that they are identified as the illegal practice
of "tying" -- that are remarkably similar to those of Microsoft in the
software market.  The article is entitled "Ecological Eugenics", and
it explains its title this way:

  [The plaintiffs are] led by Jeremy Rifkin, tireless gadfly and
  America's foremost proponent of ecological eugenics.  The new
  eugenics embraces passivity in the pursuit of supposedly superior
  genes as fervently as the old embraced intervention.

This is really confusing.  He's saying that Rifkin is a eugenicist
because he's doing the opposite of what the eugenicists used to do.
The mind reels, searches for some other explanation of what this
passage could mean, and finds none.  It's embarrassing even to explore
the alternatives.  If it even makes sense to shift the word "eugenics"
from human beings to corn, doesn't that make the people at Monsanto
into the eugenicists?  They're the ones who are trying to narrow
the world's gene pool so that only the best survive.  Rifkin is said
to "embrace passivity in the pursuit of supposedly superior genes".
What does it mean to be passive in the pursuit of something?  And
"supposedly superior genes"?  It's Monsanto that supposes that the
genes in question are superior, not Rifkin.  Or perhaps Huber means
that Rifkin is a eugenicist simply because he favors one set of genes
over another?  But then everyone involved in the whole story is a
eugenicist, including the farmers who have bred crops for thousands
of years.  Several other interpretations are possible, but none of them
makes any more sense than these.  Honestly, this is one of the most
logically muddled sentences that I can remember reading in a newspaper.

I'm not a big fan of Jeremy Rifkin, and I've always assumed that one
could say some legitimate things about him.  But maybe I was wrong,
since whatever these legitimate things might be, Huber doesn't ever
say any of them.  Remember how I said that the word "labeling" has
legitimate uses?  Huber is trying to affix a nasty label to Rifkin
whether it fits or not.  He knows that a snappy label is worth a
bushel of reason and evidence, and he is trying to manufacture a new
label, "ecological eugenics", that he hopes will work the same way as
"political correctness": even if it's not so clear what it means, it
sounds good and it lends itself to invective that is specific enough
to be persuasive but not specific enough to refute.  These sorts of
labels have been mass-manufactured in recent years, and it's high time
for someone to compile a dictionary of them.

In fact, Huber's strategy of creating primitive mental associations
actually begins in the first paragraph of the article:

  Pitched on its environmental merits, the class-action lawsuit filed
  last week against Monsanto would be thrown out in short order.  So
  the lawyers dressed it up as an antitrust case instead.  Now it's
  the Microsoft case, redirected against genes.

So even though their antitrust suit names a large and controversial
corporation, Rifkin and his followers are "against genes".  You can't
sue genes, of course, and it would be bizarre to be "against" them.
Indeed, the rational reader will be reminded that Rifkin's stated
goal is to preserve the natural diversity of genes.  Huber is working
another some/all ambiguity, playing on the plaintiffs' antipathy
toward two particular genes to stimulate a much larger association in
the primitive calculus of the lizard brain.  "Genes" here is a symbol,
and Huber is trying to capture the symbol for his side.  Those wackos
are against genes, he's saying, and we're for them.  Readers of the
Wall Street Journal's editorial page will not need much persuading
on the matter, of course, but persuading WSJ readers is not the point.
Huber's column is an intermediate product in the industrial pipeline
of twisted language, and those readers of the WSJ's editorial page who
find themselves needing some twisted language on this particular topic
will now have it ready.  Huber's sophistry will become "part of the
culture", as the PR people now like to say, sent forth to confuse the
many normal people who won't have the spare time to figure out what's
wrong with it.

Now a simple one.  When Tom Daschle objected that Republicans were
having an especially hard time confirming Bill Clinton's nominees who
happened to be black, a spokesman for Senate majority leader Trent
Lott, John Czwartacki, said:

  "Are we being asked to go back and catalogue and identify everyone's
  race, religion or ethnic background?  I mean, there's a term for
  that.  It's called racial profiling.  I thought we were all trying
  to get away from that" (NY Times 10/21/99).

Never mind the almost subliminal way that he used the padded language
of his rhetorical question to inflate the topic of racial profiling to
include religion.  The more basic problem is that racial profiling has
arisen as an issue and been condemned in settings where people were
picked out for negative state action on account of their race (e.g.,
being stopped and searched), whereas Mr. Czwartacki applies it to a
situation where people are being defended against negative state
action that is plausibly motivated by their race (i.e., failure to
properly consider their nomination).  An attempt to combat racism is
twisted to make it made to sound like an example of racism.  The trick
is to dilute the meaning of the phrase, deleting one of its elements.
It is thus also an instance of another common pattern: using "their"
phrase in a flagrantly sophistical way against them.  This strategy
has two virtues: it tends to discredits the object of the attack,
and to discredit the phrase itself by stereotyping uses of it as

It's important here that flagrant sophistry does not necessarily work
against the person who uses it.  If you do it often enough, and if
you control the press thoroughly enough that nobody ever gets a chance
to call you on it, then it's positively advantageous: it's a display
of power and a way to induce feelings of confusion and powerlessness
in your opponents.  It also illustrates the role of defaults in the
workings of power.  The powerful can discriminate all they like, and
the grounds for their decisions need never be public.  So if they can
twist the norm of social equality (e.g., "colorblindness") into a way
of suppressing all inquiries into issues involving race, including
those that hold the powerful accountable for the grounds on which they
decide things, then they win.  Mr. Czwartacki is placing himself on
the side of the cop who pulls you over for the twenty-fifth time for
no clear reason; when you complain about the obvious pattern as to who
gets this treatment and who does not, he says, "Aw, why did you have
to go and bring race into it?".

If that example does not persuade, let us consider a much worse one.
Senator John Ashcroft argued that a judicial nominee named Ronnie
White, who happens to be black, had demonstrated "a tremendous bent
toward criminal activity".  He also argued that White was against the
death penalty on the grounds that he had dissented from a particular
death sentence because the judge had "opposed affirmative action".
In reality, White had dissented from the claim that the judge in that
particular case was impartial toward the black defendant even though
one week earlier the judge had said, and I am not making this up:

  "... the Democrat party places far too much emphasis on representing
  minorities such as homosexuals, people who don't want to work and
  people with a skin that's any color but white. ...  While minorities
  need to be represented, of course, I believe the time has come
  for us to place much more emphasis and concern on the hard-working
  taxpayers in this country" (NY Times 11/9/99).

Let us take the senator's phrases in reverse order.  First, he chose
to gloss the judge's racist statement by saying that he had "opposed
affirmative action".  One might well infer that the judge most likely
opposes affirmative action.  But the judge said a lot more than that.
What we have here is another example of substitution into an opaque
context: placing into White's mouth a badly twisted version of the
proposition that he was dissenting from.  (In fact the context was
opaque twice over, but never mind.)  White wasn't dissenting from
opposition to affirmative action; he was dissenting from the idea
that people whose skins are not white are not hard-working taxpayers.
Ashcroft thus spoke falsely, but in a way that is confusing, and that
admits a certain tortured paraphrase of his meaning in the event that
anyone challenges him.  The display of power here consisted precisely
in the confidence that, even if challenged, he could get away with it.

Now let's take the other phrase: "a tremendous bent toward criminal
activity".  Notice the ambiguity.  The word "bent" here could imply
that White is overly tolerant of criminal activity.  That is false,
as White's public record demonstrates, but it is not yet an outrage.
He also accused White of being "pro-criminal".  That's approaching
an outrage, but not quite there yet.  What's an outrage is the other
meaning of the phrase, the one that employs the more commonly accepted
meaning of the word "bent": the suggestion that White would be likely
to engage in criminal activity himself.  Of course, no evidence was
presented for this idea, and certainly not enough to justify the word
"tremendous".  This would be an appalling insinuation against anyone,
but in a context where primitive racial stereotypes (e.g., black
people as lazy criminals) are being dredged back to the surface it's
completely off the deep end.  Yet every single Republican senator who
voted on the nomination accepted Ashcroft's reasoning.  That is really

Rhetorical sloppiness is everywhere, and so alas is twisted language.
But what I find most frightening about the twistedness of the new
jargon is its systematic character.  The twisted rhetorical devices
that I have been identifying are omnipresent, not only in the media
but in the everyday discourse of otherwise ordinary people.  Where do
they come from?  Well, the media outlets that publicize them are quite
clear about the intentions behind them.  Their purpose, they say, is
to "annoy liberals" and to "win arguments".  The very word "win" here
is an example of twisted language; one can win an argument logically,
in the sense that one demonstrates that one's conclusions follow from
one's premises and that the other party's does not, or one can win
interactionally, in the sense that the other party gets mad and gives
up.  People who use the rhetorical devices of the current jargon are
not interested in logic.  They are just interested in winning.

How does one get like this?  It seems clear enough.  First you must
persuade yourself that "they" are evil -- not just mistaken or stupid
but acting in infinitely bad faith.  Having done so, you can persuade
yourself that "they are really the ones who" are doing whatever evil
things you happen to be doing yourself, so that nothing you could ever
do could match their perfidy.  Irrational uses of language, in other
words, are nothing compared to what "they" do.  It's just a taste of
their own medicine.  We're just trying to annoy them, just standing
up to them.  They should stop taking themselves so seriously, learn to
laugh at themselves.  That's what the jargon's speakers say, and that,
presumably, is how they inwardly justify their outward irrationality.
It's strictly an emergency measure, they imagine, and they all intend
to become rational again once the enemy is gone.

But of course it doesn't work that way.  Once irrational patterns of
thought and language become so densely embedded in the culture, so
deeply institutionalized in what was formerly the sphere of public
reason, there is no stopping them.  Ever more extreme ideas can apply
the irrational rhetorical devices in a mechanical way, until reason is
replaced by nothing but authority and will.  Do the great majority of
qualified researchers believe in global warming?  "Political science."
Was the senator quoted in the New York Times?  "Liberal media bias."
Many people screen out inconvenient facts, of course.  But what's so
frightening here is how automatic it has become, and how a repertoire
of stock phrases has now arisen for doing it so routinely.  In this
way reality itself evaporates -- not by any conscious repudiation,
but simply by the out-of-control operation of rhetorical devices that
place inconvenient elements of reality into the mental memory-hole of
"them", the infinitely perfidious others.  Thus Rush Limbaugh (9/16/99)
suggested that the Democrats had staged a church shooting in Forth
Worth in order to promote gun control.  This is intense.  Morality
evaporates as well -- everyone praises it, but when its demands
become inconvenient, one can project the demands of morality as well
into the cognitive netherworld of "them" simply by uttering the words
"political correctness".  It becomes easy to sneer at morality -- just
say, "I guess I'm not being politically correct here, haw haw haw",
and your audience will sneer along with you.

Liberals have said and done plenty of foolish things, all of which
have been thoroughly criticized.  But the hysterical irrationality
that is daily vented upon "liberals" in the media has gone far beyond
any reference to real people.  At the end of the day, "elite liberals"
are not real people in the world.  Rather, they are part of the mental
furniture of people who have learned, through the relentless twisting
of language, to dissociate anything at all that does not comport
with the arbitrary dictates of their own will to power.  This is the
insanity at the core of authoritarianism, and it is the primal vacuum
that the longing for authority promises to fill.  In reality it is
no longer a matter of liberal versus conservatives, inasmuch as those
conservatives who still have a conscience are appalled, albeit mostly
in private, at the madness they hear pouring out of the media in their
name.  If we don't somehow snap out of it, the positive feedback loop
of imaginary enemy-making will grow extreme, and then some people will
begin to cry, Only a strong hand can save us from the evil!  Is that
because those people consciously set out to bring authoritarianism
to their country?  Not at all.  After all, the word "authoritarianism"
will long since have been redefined as something that "they" intend
for "us" (for example, Francis A. Schaeffer, A Christian Manifesto,
Crossway, 1982).  Authoritarianism starts with the genuine desire by
decent people to promote morality, strengthen the country, encourage
responsibility, heal their families, and so on.  These are decent
goals.  But then by slow insidious degrees, by the bending of means in
the service of ends, by the gradual internalization of a professionally
designed irrationality that runs deeper than anyone is aware, it
gets out of control and fastens chains upon the minds of everyone who
participates in it.  That is what is happening now.

Evil hides best when all attention is focused on another evil, real
or imagined, somewhere else.  Josef Stalin really was evil, and his
existence provided a hiding place for a complementary evil in the
political culture of the United States.  That evil is now coming
to power, now showing itself, now naming its nebulous enemies, now
publicly flouting all reason, now howling in the words of the talking
heads, now daring anyone to name it.  It is a hard thing: because
evil begins by naming evil, those who name evil fall under suspicion
themselves.  But I name it here, this authoritarian thought-pattern
now settling down on my country, and now I suppose I will see if my
country retains enough sanity to care.

Some URL's.

Bled Electronic Commerce Conference, 19-21 June 2000

Who Owns On-Line Courses?

Value-Sensitive Design

Nameless in Cyberspace: Anonymity on the Internet

looking up random words as URL's

California Engineers Report Chip Breakthrough

WTO protests

article on the valuation of Microsoft's employee stock options

Campus Wide Information Systems

Zebra Zeb-Roller pens

The New Interactivism: A Manifesto for the Information Age

legal analysis of "open access"

Echelon Watch

First Annual Conference of the Association of Internet Researchers

Attack Trees: Modeling Security Threats

The iCatch Mouse Adapter
(retrofit for the iMac's round mouse)

Media Sends Mixed Messages on Y2K

Toward a Theory of Standards

The Virtual University is (paradoxically) the University Made Concrete

DOJ's proposed conclusions of law in the Microsoft antitrust case

On-Line Courses of 1,000 Students Will Become Common, Industry Group Says

Why Is Wireless Taking So Long?

E-ZPass Privacy Policy Reviewed

The Packer / PBL / Acxiom InfoBase

Open Archives project

5 Corporations Reportedly Bid to Join Universities in Distance Education

Kosovo casualty report

Cyberspace Self-Governance: A Skeptical View From Liberal Democratic Theory

Telecommunications Policy Online

analysis of MS ruling,1151,,00.html?mail

'Spies' in Your Software?