More notes about education, design, conservatism, and cheap pens.

It's fall, and time again to circulate my how-to's.  Following this
message I will send out fresh copies of "How to help someone use a
computer" and "Advice for undergraduates considering graduate school".
Please forward them to everyone who can use them.  I'm also hoping
to reenlist you in my grandiose project of getting "Networking on
the network" into the hands of every graduate student in the world.
It is too big to distribute as a plain text message, but its URL is:

If you have read "Networking on the network", perhaps you can send a
raving testimonial about it to any graduate students you might know.

And it being fall, graduate students are looking for work.  If you
are hiring new assistant professors, or really any job in the general
vicinity of RRE's subject matter that requires a new PhD, send me your
job ad (as a plain text message, not as an attachment).  I'll send a
batch of ads to the list in perhaps the second week of December.

Are you still complaining about spam?  Contrary to the best and worst
of predictions, the spam problem has not changed appreciably since the
FTC cracked down on the online pyramid schemes.  Predictions that this
enforcement action would have no effect because the scamsters would
go offshore have not come true, but then neither have predictions that
spam would grow infinitely unless it is crushed altogether.  I don't
know if the current situation represents a stable supply of spam or
whether it represents an equilibrium between increased motivation to
supply it and increased measures to suppress it.  In any event, please
let me encourage you to resume your anti-spam activism.  Most of the
information in my "How to complain about spam" is still operational:

Although I have not committed to keeping "How to complain about spam"
up to date, let me know if you find any outright wrong information in
it and I'll fix it.  From my own experience, the major focus of the
problem right now is in Argentina, where spammers operate on a large
scale and ISPs never respond to complaints (or at least to complaints
in English).  I (and no doubt many others) have written to several
of the US ISPs that handle traffic from Argentina, but nothing has
happened that I know about.  Another focus of the problem in my own
experience is an ISP called, which has failed to suppress
some chronic spammers on its system or to respond to numerous messages
documenting the problem.

My short article on "Information technology in higher education: The
'global academic village' and intellectual standardization" has now
appeared in print, and my home page has a link to a couple of Web
versions of it, one public and one proprietary.  The public one is:

Even though I circulated an early draft of it on this list, I still
want to advertise the finished version.  It's much better, due mostly
to the impetus of the referees' comments.  It concerns a phenomenon
that I consider very important: the centralizing tendencies of the
Internet.  You've probably heard that the Internet is history's most
powerful force for decentralization, shattering hierarchies etc etc.
And that's true sometimes, for some purposes, under some conditions.
But mostly it's false, and this article explains very concisely one
of the reasons why.  Briefly, much of the economic leverage from the
Internet comes from economies of scale.  Economies of scale require
using the same information at many different points in the economy,
and this in turn requires making those many points similar enough that
the information is meaningful and useful in all of them.  In the case
of higher education, it means a much higher level of standardization.
Standardization need not be a bad thing, and it need not necessarily
lead to homogenization.  But the scenario that leads from our current
situation to a global intellectual monoculture is discouragingly easy
to describe, and I describe it in this article.

The general consensus was that my last "What I'm interested in" list
of books was too long.  So at least we've narrowed down the point
of overwhelm to somewhere between 500 and 800 books.  For what it's
worth, and by popular demand, I have gathered all of my book lists,
including some older ones never before seen on this list, into one
great huge file:

This is a very long (about 500K bytes) Web page listing nearly every
book that I have found interesting enough to write down a citation
for in the last ten years.  I do not necessarily endorse these books,
but perhaps you will find something of interest.  In fact everyone who
has scanned these lists has reported hearing about numerous relevant
books that they had never heard about before.

These immense lists of books may strike you as odd.  They ought to;
they strike me as odd.  I just keep looking for a way of working
that fits with my values and strengths, and what I've found happens
to consist largely of reading everything and noticing the patterns.
Having made a routine of keeping lists of interesting books, the
effort is really not that great.  And I find the lists themselves
interesting in all sorts of emergent ways.  For example, they make
evident what lousy systems we have for matching people with the books
they ought to be reading.  In the idealized neoclassical market, this
would not be a problem: every reader would automatically know what
book out of the entire universe of books he or she most wanted to
read at any given moment.  Lots of bad books would never be written
because their authors would know that the optimal assignment of people
to books, including all of the other books that have not yet been
written, would not provide them with an adequate audience.  In order
to appreciate fully the wonder of this hypothetically effortless
matching of people and books, it may be necessary to know a little
math, specifically a set of equations that were first written down by
Cournot in maybe the 1830's.  Or it may not.  The point is, the real
world is very different from the economists' ideal, and a quick scan
of one of my book lists -- lists that would not exist if an eccentric
college professor hadn't invested a lot of uncompensated effort over a
long period to put them together -- can provide a glimpse of just how
different it is.

I am supposed to be an expert on privacy, and in that capacity I try
gamely to follow all of the news on the subject.  But it's just about
impossible.  I can be taken by surprise by something as big as my own
bank's routine sharing of customer spending patterns with marketers.
So imagine the situation of someone who has neither the expertise nor
the time to keep track of such things.  This came home to me a couple
of months ago when, for the first time in my life, I got a membership
card at a video rental store.  I've seen plenty of rented videos in
my life, of course, but the actual renting has always been on someone
else's card.  I had agreed to write a short paper about "Enemy of the
State", however, and so I needed to watch it twenty times on video.
Marching into Blockbuster and Hollywood, I felt like I was the last
human being who needed the process of renting a video explained to me.
But when it actually came time to fill out the forms, I felt a strange
calm.  This calm, I realized, was due to the Video Privacy Protection
Act (18 USC 2710), the notoriously special-purpose American law that
prevents video rental stores from revealing who has been renting what.
In my daily life I give out at least half as much personal information
as any normal person, and 95% of the time, I've realized, I feel that
queasy feeling that I don't know where the information is going.  Now
that I was protected for once in my life by a simple, straightforward,
loophole-free privacy law, I finally understood on an emotional level
the reason for such laws, and I envied the Europeans and other halfways
civilized people who have them.

In response to my wild-eyed proposal for a reflexive approach to
information technology in higher education, one person complained that
little evidence exists to believe that online communities of practice
can be made to work on any scale.  Hearing this, I was struck all
over again by the insidiousness of the concept of cyberspace -- the
idea that the Internet is a parallel reality and not something deeply
embedded in the real world.  The thing is, I never said anything about
online communities of practice.  What I said was,

  My radical proposal that every community of practice in the world
  make its conceptual frameworks explicit, and that we use the
  Internet to organize these communities around a digital library of
  XML documents that store a vast archive of all of the cases that
  the community's members have analyzed.

In other words, every community of practice is going to be defined by
the conceptual framework it uses, and we should give every community
of practice a digital library to collect documents in which students
apply those concepts to cases.  I didn't say anything about whether
the communities of practice meet online or offline.  In fact I think
that each community of practice should find its own complicated set
of practices: some entirely online, some entirely offline, and some
that combine the two.  A community might have regional groups that
meet on Thursday nights, annual conferences that gather in a different
city each year, classes that meet one weekend a month on a university
campus and ten evenings a month online, internships where the interns
both work in physical offices and engage in structured activities in
chat rooms, and so on.  The dichotomy of online versus offline is most
unfortunate when it directs attention to two very simple points in a
very large space of design options.

Another objection to my wild-eyed proposal, however, did make sense.
I said this:

  Shallow reflections on the existing practice will give us shallow
  technologies that simply pave the cow paths, doing just what we're
  doing now only with a lot of extraneous hardware.

I shouldn't have.  "Paving the cow paths" is an epithet that is best
known from the rhetoric of reengineering guru Michael Hammer.  You
will recall that reengineering was an ignominious failure, and that
the traumatic experience of reengineering has inoculated many business
people against totalistic "throw it all out and start over again
from scratch" types of management fads.  When I complained about
shallow reflections on existing practice, it probably sounded like
I agreed with this sort of totalism, when in fact my point was subtly
different, indeed the opposite.  I think that the Internet changes
things not by creating much that is new, but by amplifying things that
already exist.  The choice is simply what you want to amplify.  There
is plenty about today's university that I would not want to amplify,
and my point is that superficial reflection will at best amplify the
whole package -- the bad along with the good.  I was going for a deep
analysis of what is really good about a university education, so that
we could amplify that and not all of the accidental stuff around it.
I probably didn't succeed, but that was the idea.

In other words, there is a sense in which paving the cow paths is
a good thing.  After all, my correspondent pointed out, Boston --
the epitome of the paved-cow-path city -- is much more liveable than
most of the designed-from-scratch cities where everything is laid out
in a sprawling grid.  You don't need a car in Boston, for one thing.
There was wisdom to the cow paths.  The point, I hope, is not that
we uncritically accept the cow paths, assuming in the classically
antirational conservative style that they must necessarily have had
more wisdom than we could have today.  Nor should we automatically
ignore the past and what it has to teach -- doing so, as we all know,
will condemn us to repeat it.  The point is that serious design, and
especially the serious redesign-in-place of institutions, requires
a kind of reflection that reaches down below the ugly surfaces to
the positive essence of the institution.  You'd better hope that the
institution has a positive essence, because your redesign is going
to amplify something.  You may have to reach to find it.  The point
is, you do not have a choice.

Here is the quote of the week:

  Several times companies have called up and said basically: "we have
  this ad campaign planned.  It is $20 million.  We have already shot
  the commercial and booked the air time.  But we have discovered our
  Web site is no good.  What should we do?"  The answer is, you are
  sort of doomed.

  -- Jakob Nielsen, NY Times, 11/2/99

You may recall in the early days of Internet hype, we all went around
saying that the Internet is not like television, and that the Internet
would make television obsolete.  But now that the revolution is here,
now that multizillions of dollars are being spent on Internet commerce
start-ups, many if not most of those dollars are going to television.
The venture capitalists who are parting with this money all share the
same theory: that brand is everything in Internet commerce, that first-
mover advantages are overwhelming, and that therefore if you do not
want to spend $20 million on a brand-building ad campaign in the 1999
holiday season then you should just give up now.

The, ahem, quality of a lot of those ads would make a good topic for
another time.  But for now I want to focus on the economics.  Should
we feel just a little chagrined that it's television where those bucks
are being spent?  Okay, you say, this is the transition: we're using
television to kill television.  But does anyone believe that?  We have
discovered something important in the five-plus years since Internet
hype got going, and that something can be summarized in three words:
economies of scale.  Just about everything that goes into a Web-based
electronic commerce business is a fixed cost that has to be distributed
across a large number of customers.  If you have two sites that cost
$10m each, and site A has 1 million customers and site B has 100,000
customers, then site A only needs to recover $10 per capita on top of
the cost of the goods, whereas site B needs to recover $100 per capita.
That's an overwhelming competitive advantage for A.  It follows that
the name of the game is gathering a large audience for your site.  But
where to get that audience?  Television networks (and movie studios)
have faced the economies-of-scale logic for decades, and as a result
they are basically giant machines for assembling audiences.  They are
still much bigger audience-assembling machines than anything that is
happening on the Internet, and in fact the biggest Internet companies
are increasingly run by people who learned the audience-assembling
trade in the traditional media -- a trend predicted right on this list
by Dan Schiller in June 1997 .
This is why it's wrong to think of the Internet as a competitor to
television.  At least insofar as it is a place where information goods
are sold, the Internet and television are part of the same system.

In my last few notes I have investigated the rhetoric of some of the
more extreme proponents of technology in higher education.  This is not
because I oppose technology in higher education, or because I support
it, but rather because I want to open up a middle ground that is based
on realism and analysis rather than on simplistic dichotomies.  Some
of this reflection was provoked by the response to the paper summary
by Rob Kling and Noriko Hara that I circulated on this list, and that
caused quite the uproar on the Internet.  I recently had an chance to
see Hara present this work at a conference, and I came to understand
one more reason why communication on this subject so often fails.

You will recall that Kling and Hara's paper is an ethnography of one
distance education class.  Although it did not (and probably could
not) come across in the brief summary, the full paper consists of
a quite thoughtful and subtle description of various things that
went wrong in the course.  When Hara presented this material at the
conference, however, one of the questioners said something like this:
"Coming as I do from the biomedical tradition, I cannot see how you can
possibly claim to generalize from one single case, and even if you did
want to generalize you should have chosen a more representative case,
or at least a more sympathetic case, such as a case where the teacher
was experienced in teaching distance education courses".  I can see
how this person perceived Hara's talk the way he did.  Someone from
an ethnographic background would know perfectly well that Kling and
Hara were not trying to generalize from a single case.  They would
know that the purpose of ethnography is not to generalize to every
case but rather to develop the concepts that are required to describe
a single case in great depth.  Someone from a biomedical background,
by contrast, is trained to generalize.  The whole point of biomedical
research is to evaluate generalizations like "treatment X helps people
with condition Y".  It's important to evaluate such generalizations,
given the emotional and pecuniary incentives that very often exist to
believe such things whether they are true or not.  From the biomedical
perspective, "X" can perfectly well be "distance learning" and "Y" can
perfectly well be "wanting to learn psychology" (or whatever).

I am pleased to report that the biomedical guy was perfectly polite
about what he evidently took to be an appalling lapse in reason and
method.  But I can easily imagine others being less polite, especially
when they are typing messages on the Internet rather than addressing
face-to-face an obviously very nice person like Noriko Hara.  This
culture clash, basically between qualitative and quantitative styles
of research, is widesrpead.  I am sorry to say that I have contributed
to it myself.  But if higher education is going to adopt advanced
information technology in an intelligent way, and not a mindless way
that is disastrous for the institution and the society, then everyone
needs to get along.  Here is a rule of thumb.  Qualitative methods are
good at questions and quantitative methods are good at answers.  Every
application of every technology has a life cycle.  Early in the life
cycle you need to do qualitative studies so you know what the questions
are, and later in the life cycle you need to do quantitative studies
so you know what the answers are.  Bad qualitative studies either fail
to ask new questions or pretend to offer new answers of a generalizing
sort.  Bad quantitative studies ask bad questions or extract answers
that their numbers don't support.  Each side is necessary to keep the
other side honest, but that won't happen if each side projects its own
presuppositions onto the other.

It being early days for most uses of information technology in higher
education, we need lots of qualitative studies.  NSF has been good
about requiring a wide variety of technical projects to include money
for anthropologists to evaluate them, and some very good work has
resulted.  No institution should be making any big public claims for
its use of technology unless it at least recruited an anthropology
grad student to hang out and interview everyone.  And when people
do make quantitative claims for the efficacy of technology-intensive
teaching methods, they should be held to the same standards of
research as anyone else -- see .
Serious discussion of these matters has hardly begun, and it is badly

A conservative society is a society of orders and classes.  It is a
culture in which people routinely defer, and are expected to defer, to
their betters.  This is what conservatism has always meant, and it is
what it still means.  But you wouldn't know it to listen to the radio,
or to read the outpourings of the hundreds of conservative pundits who
dominate the media.  That's because conservativism has better public
relations than it used to.  And the key to public relations, in the
words of the CPUSA's cultural policy, is always to lead the people by
a half-step, not getting too far in front of them.  The main problem
that conservatives face is that the lower orders have become accustomed
to challenging their betters, and have developed a whole elaborate
vocabulary with which to formulate such challenges.  This is why the
public relations directors of conservativism have organized a language
war: a very systematic project of taking back words that have gotten
out of their control.

The techniques of this war are endless, but one way to slice them is
through a distinction between two categories of epithets, transient
and permanent.  Transient epithets are words that have traditionally
been used to criticize conservatives, but are now relentlessly and
sophistically applied to liberals instead.  These in turn fall into
two categories: words traditionally used to criticize the aristocratic
rich (e.g., "elites", "authoritarian", "double standards") and words
traditionally used to criticize conservative ministers (e.g., "pious",
"sanctimonious", "indoctrination").  These epithets are transient
because it will one day be necessary to reverse their meaning.  You
will hear people say things like, "You need some kind of elites, so
you should have the ones who are moral."  In fact you already hear
people saying (if you listen to the right stations) things like, "It's
inevitable to have have some kind of indoctrination, so it should be
indoctrination into the truth".  The transient epithets have several
purposes.  They serve to portray the rebellious lower sorts and their
supporters as hypocrites.  Their sophistry tends to discredit the
epithets themselves, as well as the people they are applied to.  They
enable one to engage in projective aggression by directing negative
energy at them while telling oneself that it is they who are
directing the negative energy at us.  And they cause confusion, thus
stopping or at least slowing down any attempts to organize a response.

Permanent epithets are ones that are used to stigmatize dissent from
the judgements of one's betters.  A conservative society requires a
bottomless reservoir of these.  The permanent epithets always start
as stereotypes, constructed for example by adducing a long series of
frivolous lawsuits, foolish rebellions against authority, ridiculous
objections to stereotypes, overreactions to hate speech, or whatever.
They are always vague, so that one cannot defend oneself against them.
And they are arbitrary, being applied selectively according to double
standards.  Thus, for example, people are now labeled "victims" in a
pejorative sense of the term, but only if they happen to be victims
of conservatives.  Victims of liberals, for example those oppressed by
government regulation, are still considered victims in good standing.
The very accusation is so confusing that most people's minds turn off.
Even liberal intellectuals cannot clearly explain what the problem
is.  Another example is "political correctness", a brilliant bit of
rhetoric whose workings would require a whole treatise unto itself.
What's so brilliant is that the term is used in two different ways:
absolutely any insistence on decent behavior can be labeled as
political correctness and thereby assimilated to wild stereotypes of
Stalinist political repression, but so can absolutely any example of
decent behavior.  As a result, the very act of (say) recycling trash
or avoiding ethnic slurs -- in fact, any good deed at all -- can be
made to sound like the worst evil, but only if the person issuing the
accusation feels like doing so.  In either case, the accusation is so
vague and arbitrary that there are no grounds on which one can defend
oneself against it.

Here is an example of the arbitrariness.  A few years ago I received,
as did many others, a long anonymous letter of anti-Semitic filth.
Before descending into the usual garbage about Jews casting their
lascivious eyes on Gentile children, however, it presented a letter-
perfect speech in the idiom of contemporary conservativism, with the
Jews being cast as politically correct victims and all that.  Take any
tirade against "liberals" and simply substitute "Jews" and you would
get the first half of this letter.  As such it made no more sense,
and no less, than any of Rush Limbaugh's monologues.  Now, does that
mean that conservatives are anti-Semitic?  No.  What it does mean,
however, is that modern conservative rhetoric is completely arbitrary.
If conservatives choose not to apply their rhetoric to Jews, that is
simply a choice that they have made.  Nothing in the logic of their
discourse rules the Jews either in or out as targets of abuse, simply
because no such logic exists.  That is arbitrariness.  But it is not
sloppiness.  It is not an accident.  To the contrary, arbitrariness
is the central principle of a conservative society, which holds that
social order is best maintained by allowing the better sorts to rule
arbitrarily over the lesser sorts.

It will be argued that today's conservative movement is more complex,
and indeed it is.  It is an alliance between conservatives in the
traditional sense and a variety of other tendencies, for example the
libertarians who have somehow persuaded themselves that their agenda
of freedom will be promoted by working with authoritarians.  When the
votes are finally counted, however, the conservative movement really
is conservative.  It aims to restore a society of orders and classes.
That is how it talks, and that -- and not anything about freedom --
is the only way to make sense of the laws that it is actually passing.

Having reached my actuarial halfway point of 39 this year, my mid-life
crisis is under way.  It is a new experience: I'm old.  I have clear
memories of Walter Conkrite reporting Viet Cong body counts from the
DMZ in 1965.  Oldies stations play songs that were playing on the car
radio when I was taking driver's ed.  I sent my first e-mail on the
ARPANET when it had fewer than 200 hosts.  I know what morning was
like before leaf blowers were invented.  I can remember, back before
spelling checkers, when the English language had two distinct words,
"principal" and "principle".  I cannot imagine any healthy person
wanting to eat three-quarters of what people in my culture actually
do eat -- and I live in Los Angeles.  When I turned 30 I noticed that
people who were 40 no longer looked old; now people who are 60 no
longer look old.  This is not good.

Most amazingly of all, however, I can remember what the Internet was
like a few years ago, before the successive waves of technology mania
really got rolling, each of them sweeping away all memory of the ones
before it.  I remember that until early 1997, a prominent feature of
Internet culture was the ability of e-mail messages to get forwarded
from one person to another and thereby to propagate all around the
net.  This phenomenon motivated me to publish my how-to essay on
the design of online political action alerts in the first issue of
The Network Observer.  It later became the object of much loathing on
the part of private firms who feared that these messages would get out
of control and ruin their reputations; you may recall the questionable
strategies, circa late 1996, by which companies whose practices had
been questioned in this way -- Lexis-Nexis, for example -- began to
stereotype Internet users as people who spread rumors.

Has anybody else noticed that those days are gone?  Messages no
longer propagate widely on the Internet.  All of the hoaxes have died
out.  Nobody has ever again done anything approaching CPSR's e-mail
campaigns against Lotus Marketplace in 1990 and the Clipper chip in
1994.  Why is this?  The proximal cause of the problem, in my view,
is the tendency of new-generation mail-reading programs to mutilate
messages when you try to forward them.  They indent the forwarded
messages with >'s, they wrap them around to some arbitrary number of
text columns, and so on.  A message that gets forwarded more than once
or twice is now guaranteed to be unintelligible.  But if that's the
proximal cause, is there also a distal cause?  Here's the conspiracy
theory: companies that felt threatened by propagating e-mail messages
brought influence to bear on the companies (often the very same
companies) that produce the most widely used mail-reading programs.
Think about it: people hate those >'s, but they don't go away.  Why?

I'm not sure whether I really believe the conspiracy theory.  I don't
have any particular evidence for it.  But let's not let the phenomenon
go unremarked.  In particular, let us note that this example suggests
a way that the Internet can be regulated: if a given phenomenon of
online culture does not happen unless a critical mass of users can
participate in it, and if the phenomenon requires specific technical
capabilities, then that phenomenon can be regulated by outlawing those
capabilities, which will prevent them from appearing in mass-market
software products.

Many claims have been made for hypertext.  Serious people have told me
in serious tones that hypertext marks the end of a stultifying epoch
in which readers are coerced into reading in an artificial linear way
and the beginning of a new era of liberation when readers are empowered
to read things in any order they like.  This is, inter alia, part of
the transition from "teacher-centered" to "student-centered" learning
that you hear so much about.  (You may think that I am exaggerating
the rhetoric of the hypertext enthusiasts, but I swear that I am not.)
But I personally see little value in hypertext.  Among other things
it stultifyingly provides only a few top-down choices of the passages
one is allowed to read next, as opposed to the liberatory infinity
of choices that are offered at one's very fingertips by a book.  More
basically, I think that experience has shown that it's hard to write
much of anything in a genuinely nonlinear way.  One recurring problem
is that hypertexts, such as Web-based instruction manuals, include
lots of links into the middle of a text -- but the links jump over the
passages where the crucial words that are used in the middle of the
text were defined.  So I might click on the link for "How to create
a new directory", only to be met with "Issue a 'create' command" and
no explanation of what a command is or how to issue one.  Yes, of
course, they can fix that problem.  But the way they can fix it is
to linearize the text.  I'm not saying that we should ban hypertext,
but we should get a sense of proportion.  Hypertext is a tool, and
the question of what people will make of it can only be answered in a
particular context.

Apple lives and dies on industrial design.  So naturally I was looking
forward to using one of the new iMacs.  Imagine my surprise when I
did.  As I've explained here before, the great danger of industrial
design is that the industrial designers tend to be evaluated by their
peers on what their designs look like, rather than on what it is like
to live with them.  That's why designers have been giving one another
prizes for decades for designing chairs that nobody could ever really
sit in (see Galen Cranz, Chair: Rethinking Culture, Body, and Design,
Norton, 1998).  The iMacs that I've used, sad to say, suffer some
clear examples of this problem.  I think they look ridiculous, but
I'm not an artist so what do I know.  What I do know is that *it makes
no sense to have a round mouse*.  Why?  If it's round then you can't
orient it without looking at it!  And if you can't orient the mouse
on the desktop then the mouse cursor on the screen will move in some
unpredictable relation to the direction that your hand is moving.
Really now.  And then there's the miniscule Escape key on the iMac
keyboard.  Lots of programs, Emacs for example, require that one use
the Escape key constantly.  But it's not possible to hit the iMac's
Escape key without slowing way down and pressing precisely, so that
one really enters an Escape and not one of the neighboring keys.  I'm
as impressed as anyone with Apple's doubled market share, which is an
incredible indictment of the last couple of managements there, but I'm
not buying any stock yet.

Here's my scenario for the end of Moore's Law.  Let us stipulate that
no technology barriers exist to the miniaturization of computer chips,
and that it will be technically possible to build microprocessors with
brain-like numbers of connections on something like the schedule that
seeming lunatics like Ray Kurzweil predict.  The question is whether
the world can digest those chips.  Even as we shout from the rooftops
the exponential increase in the numbers of transistors on a chip, we
don't give nearly as much press to a countervailing phenomenon: the
exponential growth of the fixed costs of producing those chips.  A new
microprocessor fabrication plant costs something like $2 billion to
construct and has a useful lifetime of perhaps 18 months.  That money
is spent before the first chip is made, and whole fixed cost must be
distributed across all of the people who buy the chips.  If the trend
continues then fab plants will soon enough cost trillions instead of

Those costs will not be recovered unless we can find several orders of
magnitude more uses for high-powered computer chips than we have now.
Yet many people are perfectly happy with 250MHz processors, and in
fact I just got done paying money for a Macintosh Powerbook 2400 with
an 180MHz processor because it was powerful enough for my purposes and
a lot more portable and less doofy-looking than an iBook.  That's a
real problem.  Of course, once we figure out how to get high-bandwidth
data connections out to 100M homes, we will have a mass market for a
lot of computation-intensive video applications.  George Gilder's 1990
scenarios of general-purpose digital machines replacing television
sets might even happen.  But that is not going to happen soon, and
even then we are only talking about low numbers of billions of chips.
So my scenario is that the exponential growth of processor power is
self-limiting because of sub-exponential growth in demand for high-end
processing.  Other scenarios are possible, of course.  We might keep
on inventing exponentially many uses for the exponentially greater
processor power.  Or new fabrication technologies might be discovered
that require sub-exponential growth in fixed fabrication costs.  Place
your bets.

You've probably heard by now that the poor sad Los Angeles Times, once
an adequate newspaper known for its thoughtful features, is being run
by extremely stupid people who do not understand what a newspaper is.
Their supreme leader is literally a former breakfast cereal executive.
Now he has hired as LA Times publisher the stunningly clueless Kathryn
M. Downing, who was recently discovered to have countenanced a deal
whereby the newspaper devoted a whole obsequious issue of the LA Times
Magazine to the new Staples Center and then split the profits on the
issue with (!) the new Staples Center.  Former publisher Otis Chandler
broke a long silence to denounce this amazing bargain, to which Downing
responded as follows:

  Otis Chandler is angry and bitter and he is doing a great disservice
  to this paper.  And that's too bad because when he was publisher, he
  did wonderful things.

As part of my new campaign to confront professionally twisted uses of
language, I want to draw attention to the first six words of Downing's
quote: "Otis Chandler is angry and bitter".  Notice how they play with
English grammar.  They admit two readings: (1) "Otis Chandler is angry
and bitter about the mess that his successors have made of his life's
work" and (2) "Otis Chandler is an angry and bitter man, independently
of any objective reality".  The quote "works" by insinuating (2) while
not outright denying the truth, i.e., (1).  It is a special case of
the subrational manipulation of language that I have been describing
in recent messages.  I call it "separation": breaking the association
between two concepts, a cause (the behavior of the new management)
and an effect (OC's feelings), in such a way that the effect is made
to seem like a perverse inner state of the individual with no real
cause, and doing so covertly through ambiguous language whose agenda
is obvious but can still be denied.

This trick, it should be noted, is distinct from another common trick
that on the surface seems related: applying the word "disgruntled"
to a fired employee who issues accusations against his or her former
employer.  In that case, separation of (claimed) cause and effect is
only the first step, the second step being the use of ambiguity to
deniably insinuate a new cause.

For now we'll just name these phenomena.  But soon their deeper logic
will become apparent.

The world consists of two kinds of people: people who understand the
point of writing about cheap pens, and people who do not.  The people
in the latter category are strange, but they're allowed on the list
anyway.  Be nice to them.

Speaking of cheap pens, I want to thank Mark Warschauer for sending
me some pleasantly wide-bodied gift-shop pens from Cairo -- they're
covered with Egyptian tomb inscriptions and prominently marked "Micro
Ceramic Pen Ares 0.5 Korea", a cultural combination that will raise no
eyebrows in Los Angeles -- and Frank Ritter for sending me a Papermate
Gel-Writer (which had already been in the mail when I found it on
my own) and a Penn State pen that was unusually good for a "stick"
ballpoint.  Thanks also to Henry Lieberman for reminding me about
the Micron Pigma, a plastic-tip pen that comes in sizes from .05mm to
.7mm.  I used Pigmas for a while when I was first experimenting with
cheap pens several years ago, but they're not for me.  You have to
write very gently with them or you'll mash the point.  I do find them
useful, however, as a reminder that people can be different from me.

In Sofia the other day I paid a street vendor a couple of new levs for
a (take a deep breath) Beifa Free Ink Roller 0.5 Be-A3, which declares
that it is made of "Materials from Germany.Swiss", whatever that means.
Boy is this pen a loser.  It's a cylindrical-tip liquid-ink pen of
more or less the standard design, except that writing with it is like
trying to write with an X-Acto knife.  You know how the X-Acto knife
keeps deciding that it wants to go off and cut a nice straight line
in some different direction from the one you had in mind?  That's what
the Beifa is like.  I find myself going back to redraw at least one
letter in every other word, but then the pen is delivering so much ink
that the newly redrawn letter gets swollen and the whole thing looks
like a mess, yet a somehow a very precisely drawn mess.  I bought this
thing about a block from the old Communist Party headquarters, a truly
strange building that looks like an aircraft carrier cutting down the
middle of a city street.  I gotta say I feel for the people in Sofia.

In case you're wondering, when I end up with a loser of a pen, I don't
throw it out.  Sometimes I give it to someone who is not a cheap pen
connoisseur, but usually I keep writing with it until it's done, which
with a lot of these pens isn't such a long time.

We're still looking for a source for the recommended Zebra Zeb-Roller
DX7.  We've had a sighting at Staples, but I haven't been able to
confirm it.  If you locate a source of Zeb-Rollers, do let me know.

Finally, the mystery of the advanced gel pen technology is starting to
affect productivity.  Despite what you may have heard, the technology
probably does not come from extraterrestrials.  Can someone please do
a patent search?  It makes sense to start with Pilot, whose G-1 is an
excellent example of whatever this technology is.  The main question,
indeed the central design question for all of these cheap pens, is how
the ink gets chased down toward the point as you write with it.  Does
it involve a vacuum, or what?

Have a look at .  I do
not know if this person's claims are true, and I am mentioning his
URL with some misgivings.  His claim is essentially that if Microsoft
were to account for all of the stock options it has distributed to
its employees then it would be insolvent.  He writes like a crackpot,
so maybe he is one.  I don't know.  What's interesting is why his
claim is so plausible in the first place.  You might be aware of a
battle that has been going on for some time now about the accounting
standards for stock options.  It is an obscure matter, to be sure,
but it is crucial for computer companies, many of whose books would
look very bad indeed if they had to put the value of those options --
which can be priced perfectly well -- down on their balance sheets as
current compensation to their employees.  Some accountants have been
proposing that firms be required to do just that, and the computer
industry has been working very hard to prevent any such rule from
being adopted.  As anybody who reads the business pages carefully
knows, there is a book to be written about the financial history of
the computer industry -- in fact, AOL would be an even better starting
place than Microsoft.  Maybe that will happen now that a judge has
made it official that computer companies have to obey the law like
everyone else.  At which point it will be possible to evaluate quite a
few interesting claims.

Some URL's.

The Radical Impossibility of Teaching

article on universities "leveraging their brands",1449,7122,00.html?01

proceedings of Int'l Conference on Privacy and Personal Data Protection

Are you a better driver? Insurer's device proves it

Networks for People Conference and Digital Divide Summit

UNext ("we deliver ... superior human capital")

University Access

Proceedings of Int'l Conference on Privacy and Personal Data Protection

Social and Economic Implications of IT Bibliographic Data Base

How the Internet Ruined San Francisco