T H E N E T W O R K O B S E R V E R
VOLUME 3, NUMBER 5 MAY 1996
"You have to organize, organize, organize, and build and
build, and train and train, so that there is a permanent,
vibrant structure of which people can be part."
-- Ralph Reed, Christian Coalition
This month: Rethinking hyperlinking
Libraries and collective cognition
Advice for liberals
Welcome to TNO 3(5).
This issue is a collection of footnotes on topics that will be
familiar to regular readers of TNO. The first article concerns
the design of Web pages. A major problem now is that it's hard
to know what you'll get when you click on a hyperlink. I think
the whole model of hypertext is misguided or at least inadequate:
just underlining or highlighting a word doesn't tell me what's
behind it. The key is stable, well-designed genre conventions.
Next I continue last month's discussion of digital libraries.
This month's version of the story overlaps with last month's,
but it cuts through the issues from a different direction. It's
important to think about "digital libraries" as something beyond
great masses of digital documents with a user interface on them.
Instead, let's think about the diverse ways in which networked
computing and professional librarians can support the diverse
ways that communities think together. I'm not a librarian, so I
have limited authority to pronounce on these matters. What I am
is a heavy user of libraries who wants to ensure that the world's
highly evolved library system is not ruined in a generation by
the willy-nilly imposition of simplistic models of information
derived from computer science.
This month I have also assembled some advice for liberals. My
theory is that every social movement develops blind spots once
its leaders have established a steady income for themselves.
The problem is that these leaders' whole identity is defined in
terms of the battles they won in the past, not the battles they
face in the present. As a result, they literally cannot see the
new movements that arise from the grievances that their success
has created. In any case, *something* has to explain the dearth
of cogent refutation by liberal spokespeople of the detailed
arguments of the ascendant conservative movement. Right now we
seem to be at a lull, as Bill Clinton shape-shifts himself into
the center and the conservatives stand about speechless at the
destruction caused by Pat Buchanan's primary campaign. But the
fundamentals are unchanged. Maybe the liberals will take the
opportunity to rebuild, and maybe they won't.
Once again I am collecting arguments against privacy for a future
issue of TNO, perhaps in July. The most common argument against
privacy that I have omitted in previous issues, it would seem, is
"you needn't worry if you have nothing to hide". This argument
is deeper than it looks, as I'll explain in the next round-up.
In the meantime, please send me any arguments against privacy
that you encounter. I am particularly interested in arguments
that pertain to specific issues such as medical or financial
privacy. I have concentrated primarily on *bad* arguments
against privacy, but if I find any good ones then I'll mention
those as well.
A footnote. This month I happened to see the last half-hour
or so of "Johnny Mnemonic". Skip forward if you don't want
to know how it ends. In the climactic scene, the leader of an
underground organization broadcasts the cure for an AIDS-like
disease, which a global pharmaceutical company has kept secret
because it can make more money by treating people as they die.
Preparing to bounce this video signal off a satellite by means
of jury-rigged machinery controlled by a telepathic dolphin (of
course), he prefaces the broadcast by telling everyone to turn
on their VCR's and explaining that the company hadn't wanted the
information to get out. Despite my hardened shell of cynicism,
part of me felt a ridiculous rush of identification with this guy,
as if I were Fighting The Power by running a large global mailing
list. It's an excellent fantasy: the technoanarchist masses are
out there, counter-hegemonic VCR machines at the ready, waiting
to be activated by the right information. No need to listen,
no need to converse, no need to write articles or direct films
that people can understand, no need to travel into neighborhoods
where you can't get good coffee -- no need to do anything but
uplink video. The film pulls off this fantasy in part through
the character of the underground leader, a hip-hop DJ type with
strange facial markings. His blackness is employed as a sign
of political authenticity, even as hip-hop's appropriation of
technology is transformed from an intricate cultural dialogue
to a broadcast model that encourages the heroic nerd without
requiring this nerd to go outside. E-mail and video are indeed
important, and information is important too. But they have to
be part of something.
The WorldWide Web, for all its flash and color, still resembles
the early days of radio: the sound quality may have been crackly
and the reception intermittent, but much of the pleasure came
from the very idea of it. We look at the Web through special
glasses that let us neglect the reality in favor of the fantasy
of what it will someday become. When this effect wears off, we
will suddenly start asking whether those Web pages are actually
useful for something. A Web page can be cool in complete
isolation, but it can only be useful in the fullness of mundane
practical reality. Right now we don't mind waiting fifteen
seconds to follow a link because we think of it as a one-time
thing. But once we get back down to business we'll be waiting
fifteen seconds once per day, or once per phone call, or once
per page of an electronic magazine. Then we'll mind.
When that day comes, I think the Web will have to change a great
deal. I would like to suggest that we will want to change some
basic categories of the whole hypertext paradigm. Let's consider
the hyperlink -- the colored or underlined bit of text that you
can click on to jump to another Web page. My big problem with
hyperlinks is that I rarely have a clear sense of what I will
get if I click on them. That's okay when the ruling metaphor is
still "browsing" or "exploring", so that getting lost is normal
-- see TNO 1(10). But it's not okay when I have a concrete goal
and no extra time for Webspace tourism.
How can Web page designers induce accurate expectations about
their hyperlinks? One obvious approach is to explain, with a
brief bit of text, perhaps in the margin or on a bullet, what
will be found at the other end. And that's good. Web browsers
might be able to help in limited ways as well. Ben Shneiderman
and others have suggested that a browser display several lines of
information for each hyperlink, the way that the URL now appears
on the bottom of the screen in some Web browsers when the mouse
passes over the hyperlink. And that's good too. But I want to
back up and consider the question more fundamentally.
Consider the front page of the Wall Street Journal. It's
tremendously well-designed. It has a stable format from day to
day. It has six columns, and they always use the six columns
in exactly the same way. You always know where to find the
feature articles, the humor article, the brief news items, the
statistical graph (top center), the summary of articles inside
the paper, and so on. And we shouldn't forget the more mundane
conventions that other newspapers share: you always know where
to find the date, the page numbers, and so on.
The WSJ's front page, then, supports expectations by always
putting the same stuff in the same place. But beyond that, it
also supports procedures for finding what you want. Business
people value their time, and many of them have routines for
reading the paper. One common routine is to scan the column that
summarizes company reports. The company names are thoughtfully
listed in bold type so the reader's eye can catch on companies
that he or she does business with, competes with, wants things
from, might look for a job at, or whatever. Another routine
is just to read the headlines. Of course, the readers of many
newspapers scan headlines -- that's why they're there. But
the WSJ provides several different headlines, in different type,
summarizing the story in several different ways, but always with
the same sort of "rhythm".
The WSJ's front page includes many hyperlinks: the pointers
to inside pages where continuations of stories can be found.
Stories that start on the first page of a given section always
continue inside that section, and the WSJ, unlike many other
papers, tries not to string a long article across several pages
to encourage you to notice as much advertising as possible. The
design conventions create strong expectations about what sort of
thing you'll find if you follow each link to the specified page.
We can think of these things in terms of "good design". But some
conceptions of good design are better than others. We should
ask, "good design for whom?" and "good design for what purpose?".
The WSJ exhibits good design for people who read it frequently
and as quickly as possible. It exhibits average design for
people who are reading it for the first time or who are just
browsing and exploring. It even embodies particular ideas about
what the reader's interests might be -- for example, an interest
in tracking particular companies.
As regular readers of TNO will have anticipated, I find it useful
to treat the WSJ's front page not as a singular artifact but as
a genre. When you buy an artifact like a chair, you can use the
same chair repeatedly. Artifacts whose main purpose is to embody
information, though, are different: much of their value goes away
after you use them once. But your life is probably organized
in cycles. You will probably need other informational artifacts
of the same general sort in the future, and so you will probably
obtain a steady stream of them from the same source. This, as I
have explained in TNO 2(11), is the purpose of designing genres.
The WSJ's front page is a genre in this sense. Each of article,
long or short or tiny, also instantiates a genre in its own
right. The overall structure of the newspaper, and the internal
structure of each article, fit into certain kinds of activities.
This "fit" has several aspects, one for each aspect of the
embodied activity of reading the newspaper: goals, institutional
relationships, perceptual affordances, properties of the paper,
arms and hands, bottom lines versus details, reportage versus
opinion, and so on.
On this analysis, the key to good hyperlinking is not necessarily
a detailed summary, though summaries are often useful. The
key to good hyperlinking is design that fits into the reader's
routine activities. If the genre includes stable relationships
among its parts then the reader can form stable expectations
and incorporate these expectations into routines.
The problem with Web pages is that they do not have enough design
conventions. Except in unusual cases, a successful Web site will
need to fit into a user's routine ways of life, either creating a
new set of stable expectations or drawing upon expectations that
users bring from their interactions with other Web sites or other
media. This means that Web site designers need to ask themselves
many questions about their users and their users' lives. I've
listed some of these questions in TNO 2(11). One such question
is whether the materials are intended for users who use such
materials regularly as part of a routine, or for users who will
only use those materials once.
Many other problems will have to get sorted out along the way.
For example right now it's too hard to tell how a Web page has
changed since the last time you visited it. Web pages that are
organized as regular publications can appear in entirely fresh
"issues" on a regular basis. But what about the nebulous middle
ground of "living documents" that just change whenever the
designer wants to change them? If the Web server knows who you
are then maybe it can synthesize a page that reflects what has
changed since the last time you visited. Or maybe not. In any
case, we need better conventions than **NEW!** icons on bulleted
hyperlinks. I don't think that these problems can be solved in
a completely general way. Each solution will depend on the uses
people have for the pages and the genres that make pages useful
by supporting stable expectations about them.
Libraries and communities.
The Internet allows us to define library work on two distinct
levels: the individual patron and the patron community.
"Community" is obviously a heavily loaded term, but here it
refers to a set of people who occupy analogous locations in
society -- or, put simply, people who have something important
in common. This approach has several virtues. It encourages
us to take a person's social identities and roles into account
when analyzing their information needs. It reminds us that the
center of mass of people's lives is located somewhere outside the
library, in the relatively stable pattern of relationships within
which they negotiate their way in the world. And it provides us
a way of imagining a collective patron: an active, self-conscious
social group, whether formally organized or not, as the
beneficiary of librarians' assistance.
The hard part is translating these concerns into design. Once
librarians do conceptualize their patrons' needs in this broader
way, how can they shape the evolution of digital libraries
in order to support the broadest conception of the work of
librarians and patrons alike? Do the necessary design strategies
have consequences for basic architectural decisions -- including
decisions that might already have been made for some systems?
And how can these strategies be translated into a technical
agenda and argued for in ways that win arguments in a technical
domain? These questions certainly seem important, and it is
quite possible that they are *urgent* as well.
To answer these questions, it would seem important to know more
about how communities use information. In particular it would
seem important to develop analytical categories that let us
talk about the uses of information in community members' lives.
A variety of existing intellectual traditions provide useful
hints in this direction, ranging from cultural and linguistic
anthropology to symbolic interactionism and ethnomethodology to
activity theory. To provide some concrete sense of what it would
be like to open up these questions in a powerful way, I want
to briefly discuss a particular theoretical trend, namely genre
theory. A genre, as we all know, is a relatively stable and
expectable form of communication. Genres of documents such as
research articles, subpoenas, menus, and Interstate Highway signs
embody complex relationships between the people who create them
and the people who use them.
In particular, these genres embody certain *strategies* of their
creators, and they *fit* into certain *activities* of their
users. The activities of writing and reading research articles,
for example -- I take the case from Chuck Bazerman's book
"Shaping Written Knowledge" -- are typified and orderly; they
fit into the social system and the daily routines and the career
trajectories of the research community. When a researcher "needs
information", the "thing" that they need isn't just "information"
as a generic stuff. Rather, their "need" is something defined
within the categories of that particular community. These
"needs" are qualitatively different -- defined in different terms
-- from the needs that come up in the institutionally organized
lives of stamp collectors or political pundits or high school
biology teachers or entrepreneurs.
It seems to me that a central opportunity and challenge of
digital librarianship is comprehending, valuing, and supporting
the qualitatively distinct kinds of needs that arise within
different communities. How can tools be fashioned that are
responsive to, for example, an entrepreneur's needs -- tools
whose design allows them to fit into the practical activities
of actual entrepreneurs' lives? And I think it's important to
*frame* this question in a broad enough way. It is a commonplace
that a networked society allows for the strengthening of ties
based on commonality of interest, so that "communities" in our
sense of the term can engage in collective thinking, collective
working, collective organizing, and both old and new forms of
mutual assistance. The tools that digital librarians can provide
will only be useful if they are useful in *that* context -- the
context of a *community's* collective life as organized through
mailing lists, conferences, newsletters, social networks,
traveling consultants, issue campaigns, rumors, parties, and
These might seem like great demands -- adapting the tools of
digital information access to the qualitatively diverse needs
of thousands upon thousands of different communities, each with
its own categories and customs and geographic distributions and
skills and typified forms of activity. The good news is that
a networked society also provides the means to *address* these
diverse needs. The future of librarianship, I want to suggest,
involves *the collective work of networked librarians to support
the collective lives of networked communities*. Digital library
systems will enable this radically new style of librarianship if
they support three crucial functionalities:
* building indexing and retrieval tools that can take account of
the meaningful formal features of diverse genres of materials
in diverse media
* permitting distributed groups of library professionals to
develop tools for particular communities by building upon
a common substrate through the assembly of object-oriented
* integrating digital information resources with the collective
lives of communities, for example through mailing lists and
community-specific computational resources maintained by
professional societies and the like
Librarians have always adapted their methods and resources to
the needs of a diverse patron community. The networked society,
though, both enables and requires us to throw this active
adaptation to diversity into high gear. A significant danger,
I believe, is that emerging digital library architectures will
embody a one-size-fits-all philosophy, or else analysis on the
level only of tailoring of interfaces to the preferences of
individuals. Only once we study and appreciate the profound
qualitative diversity of communities' collective lives, genres
of documents, and occasions for using these documents in
routine typified activities will we be in a position to design
information system architectures that are truly responsive to
Advice for liberals.
My liberal friends need a lot of help right now. Many of them
are living in the past, fighting a different decade's fights.
Getting hammered has its advantages, though, and one of them is
that it clears the ground for rebuilding. Much of the problem
is simply realizing that there's a problem, and that's the
starting point for some advice that I've been formulating and
reformulating for years now of reading and talking to people:
(1) Wake up. Something is happening and you don't know what it
is. Accept that you are getting hosed, and that things are
going to get much worse from your perspective before they
get much better. Forget about implementing any more of your
policy agenda for the foreseeable future. Concentrate on
defending what's worth defending, abandoning the rest, and
getting used to being the opposition. Yes, the polls look
good for Bill Clinton now. But he's selling you down the
river. And presidential polls have little to do with the
district-by-district shifts that determine the structure of
Congress. You simply must face the fact that you are losing
the South, which is at least one quarter of your coalition.
(2) Read conservative publications, lots of them, and regularly.
Don't just shake your head and say "we know what's wrong
with that", because I'll bet that you actually don't. You
might start with Friedrich Hayek's "The Road to Serfdom".
Subscribe to Policy Review, Christian America, and The
Standard. Read an anthology of P. J. O'Rourke's humor
columns and get used to the idea that you are being mocked.
When you're ready for the really heavy stuff, read a journal
called "Heterodoxy", which focuses on the academic left. You
will feel a powerful temptation to blow these people off as
nuts. Resist -- what matters is whether the ideas are useful
in assembling a new coalition or fragmenting yours. Instead,
internalize the arguments, admit to the grains of truth they
often contain, and learn not to present such a big target.
(3) Get beyond identity politics. Theories of marginality
really are self-fulfilling prophesies. Some people really
are divisive. A focus on words and symbols, beyond a
certain point, really does draw your attention away from
the actual *things*. Restore your focus on the skills
and processes of empowerment, articulating ideologies that
forge inclusive coalitions based on democratic values. You
may think you're doing this, but you're not. Attack your
enemies, not your friends.
(4) Refuse to be stereotyped. Pay attention to the explosion
of new stereotypes aimed at your allies ("environmental
wackos", for example). Constantly enumerate the stereotypes
and identify them for what they are. Note that they are not
principally the racial and sexual stereotypes of olden days.
Assertively use the media to teach people to recognize the
rhetorical devices of stereotyping and the forms of public
relations work that depend on stereotyping, and use real,
(5) Do vastly more opposition research. The fact is, the right
has a better mastery of the art of politics than you. Stop
dismissing them as stupid, backward, oblivious, hate-driven,
or conspiratorial. Some people do, of course, have any
combination of those qualities. But you'll blind yourself
so long as you stereotype your whole opposition in those
terms. The fact is, they understand you a whole lot better
than you understand them. The main reason they are winning
is that they have read you and listened to you and found real
weaknesses in your accustomed modes of argument. You need
to do the same for them. This will take time, and it will
not be fun.
(6) Stop surrendering powerful words such as family, nation,
truth, science, tradition, religion, merit, responsibility,
and character. If your opponents have given these words
false meanings, persistently restore their true meanings.
That is to say, contest the signifiers. Use the words!
Forget the whole strategy of the counterculture; be the
culture instead. Repeatedly say things like: children
need families; crime is bad; our country has many positive
traditions; science and technology contribute much that
is beneficial; people are responsible for what they do.
Use the words right and wrong, true and false, good and bad.
(7) Train many more activists. Teach them organizing skills,
media skills, basic listening skills, and coalition-building.
Build a much better infrastructure to support your activists,
connect them to one another, and supply them with tactics and
arguments. Build networks to recruit activists, especially
in churches and on college campuses, and then use these
networks to help activists develop their skills and projects.
If you're already doing this, do much more of it. Yes, this
will probably require you to divert resources from putting
out fires in the short term, but you'll have no future unless
you invest in it now.
(8) Build your network of intellectuals. Insist that your think
tank people produce their materials in a form that activists
can use -- no more long reports that sit on shelves. Direct
funding to those organizations that formulate clear, useful,
tested messages backed up by a steady stream of reliable
facts -- and reevaluate funding for the others. Sustain
frequent, substantive contacts, based on equality, between
activists and sympathetic academics. Tell the academics
who express themselves in flamboyant intellectual codes to
imagine what they'll sound like quoted selectively in the
newspaper. Build and sustain new media that provide channels
of communication between intellectuals and activists, and
among activists themselves.
(9) Read the Bible if you don't already. Pretend that it's the
Middle Ages and that all political arguments must be couched
in theological terms. Maybe that's not quite true yet, but
by the time you catch up it will be. Commence historical
study of the original transition to a secular culture, to
help everyone recall why this once seemed like a good idea.
Don't beat up on religion -- beat up on political abuse of
religion. When the right presents Bosnia as an example of
multiculturalism, present both Bosnia and Northern Ireland
as examples of religion in the public square. Spirituality
is primarily a liberal force, even when -- especially when --
it is coupled with a strong ethic of personal responsibility.
(10) Learn and teach logical argument, clear writing, and critical
awareness of grammar and rhetoric. Forget the postmodernism.
Draw public attention to illogical arguments right away,
and I mean instantly. Don't just let them slide. Once they
become part of the culture, it will take two generations to
The street numbers of homes and businesses are usually not very
well marked. Looking for a given street address while driving
a car is therefore frequently dangerous. This may seem like a
petty thing, but it really bothers me and it's interesting (it
turns out) to think about. Consequently, I wish for a device in
my car, perhaps attached to the dashboard just above the steering
wheel, that tells me which street numbers can be found to my left
and right. This device might use some sort of simple digital
wireless protocol to talk to other devices mounted alongside the
street, or else it could use GPS or some other positioning scheme
together with a digital map of streets and numbers. In a simple
form, the display could look like this:
| | |
| 319 | 320 |
| V |
meaning "number 319 is to your left, number 320 is to your right,
and the street numbers are decreasing in the direction that you
are driving". More elaborate displays could include additional
information such as the street name, cross streets in front and
behind, current compass heading, speed limit, and distance from
a desired destination.
Why are street numbers so hard to see? Maybe because visible
street numbers are a public good. Having provided visible
street numbers to one passer-by, you've made your street numbers
visible to everyone. People can use your street numbers to find
addresses besides your own. And if my house is number 207, it
is only in my interest to make my number visible enough for my
own guests to find, and not for people who are looking for other
houses on the street. In many cases these may be the same,
but in other cases they may not be. Few shopping centers, for
example, seem interested in helping people find them by street
address. And some people just don't have guests, or they give
directions by the color of the house and expressions like "third
house on the left with the Chevy on blocks in the driveway".
Another problem is that street numbers are most often attached
to houses by builders, who sell houses based on how they look
to people who haven't bought them yet. Such people usually have
much bigger problems than discerning whether the street numbers
will be visible from the street at night. Much better if the
street numbers aren't too obtrusive, marring the overall vista of
the house. Then once the house is purchased, more things happen.
Plants often grow up covering the street numbers. Guests to the
house may be reticent to complain about the trouble they had in
finding it. And even if one house does get more visible street
numbers, it will probably necessarily break the convention on
that street for where those numbers are located and what they
For all of these reasons, visible street numbers really should be
provided collectively. The town I grew up in actually paid for
street numbers to be painted on the curb in front of every house,
on the grounds that drivers of emergency vehicles needed to find
them quickly. But this only worked because cars were not parked
densely along the curbs in most parts of that town. Would it
be worth setting up a street number grid in the wireless ether
so that emergency vehicles can find storefronts? Would delivery
services like FedEx be willing to pay part of the bill, provided
other other competing services don't get a free ride? And then
maybe companies wanting to offer value-added services based on
that grid would pay some percentage to maintain or extend it.
A larger theme is that the whole world is becoming encased in all
kinds of digital representations. Every person, place, and thing
of any interest to any powerful organization is developing what
I've called a "digital shadow" that tracks it in real time. The
maintenance of digital shadows ultimately requires a coordinate
system to be laid out upon the whole world. That is what GPS
is about, but it is also happening in thousands of other ways
as well. Perhaps all of those thousands of representational
practices will converge and we will all be knitted into a
global representational grid. This will have many advantages,
particularly when regular people consume representations, as
in my thought-example here. But it will have many disadvantages
as well, particularly when others consume representations of us.
We do have time to make choices about this prospect.
This month's recommendations.
Mary Karr, The Liars' Club: A Memoir, New York: Penguin, 1996.
The hype when this book first appeared was deafening, so I waited
for the paperback to read it. I read it just in time, it turns
out, because the hype has gotten even louder. The New York Times
Sunday Magazine, for example, just reported that "the age of the
literary memoir is now" and presented Mary Karr as Exhibit A.
For once, I can report, the hype is right. (I suppose that means
that it isn't even hype, if "hype" is short for "hyperbole" as
I have always supposed.) "The Liar's Club" is the true story
of Mary Karr's spectacularly dysfunctional childhood in a swampy
Nowheresville in east Texas. Her stories are so amazing that
I don't think they would work as fiction. She recounts with
shattering clarity the perceptions and feelings of a little girl
trying to figure out why her mother is so crazy. It's all in
east Texas dialect, with sentences such as "I shit you not" that
produce great incongruity when she offers her occasional comment
in the voice of the grown-up writing teacher that she has somehow
become. And a fine teacher I imagine she must be.
Sara Diamond, Roads to Dominion: Right-Wing Movements and
Political Power in the United States, New York: Guilford Press,
1995. American liberals who find it convenient to imagine the
right as a batch of marginal nutcases or a monolithic machinery
of fascists will benefit from this discerning history of
right-wing social movements and their relationship with the
government. Diamond traces the emergence in the 1950's of the
"fusionist" strategy that somehow unites a powerful coalition
of authoritarian and libertarian conservatives today, while
also tracing the formation and decline of the more extreme racist
branches of the right. Her book is no substitute for reading the
original texts and ongoing published debates in the conservative
movement, but it does provide some necessary context for them.
Gordon Cook, National Information Infrastructure: The Dark Side
in Washington State. This is a special edition of a newsletter
called The Cook Report on Internet-NREN, which is perhaps the
best independent voice on directions and implications of Internet
architecture. This particular issue is, to my knowledge, the
only study of the full range of privacy-threatening technologies
under development in a single geographic area. Although the
details are specific to Washington State, most of the trends that
Cook finds there can also be found throughout the United States.
Cook's newsletter is one of those high-price, low-circulation
publications that everyone hopes will become viable on a lower-
priced, higher-circulation basis once such publications can be
distributed effectively on the Internet. Which is to say, it's
not cheap. The writing is not exactly polished, and much of the
book-length report is taken verbatim from dozens of interviews
that he conducted over a few weeks. Still, it's a unique and
important accomplishment. For details on ordering the report or
subscribing to the newsletter, see the Cook Report Web page at
Geoffrey A. Moore, Crossing the Chasm: Marketing and Selling
High-Tech Products to Mainstream Customers, New York: Harper
Business, 1991. I adore this whole genre -- book-length business
cards by management consultants conveying their astounding vision
and what they'll do for you once you hire them. They're written
by extremely persuasive people who have had a lot of experience
refining their sales pitches, and as a result they are incredibly
compelling. When I get done reading one of them, I feel this
appalling urge to get myself out there and conquer some markets.
They have an interesting property that sociologists have noticed
in more organized professions: the knowledge they contain is
formalized just enough to be convincing and comprehensible, but
it's not formalized enough that simply reading the book actually
qualifies you to do any of it. (See, for example, Andrew Abbott,
The System of Professions, University of Chicago Press, 1988.)
Moore's particular book is concerned with the life cycle of a
high-technology product. He observes that many companies fail
to grow beyond their first few customers. The reason, he argues,
is that high-tech markets are divided into innovators (the people
who love gadgets and will buy anything cool), early adopters (the
people who are strongly committed to a vision and are willing to
buy and unproven product that supports that vision), the early
majority (the people who want new stuff but want it to come in a
safe, reliable package), the late majority (the people who don't
want to buy new stuff, but will buy it anyway once everyone else
has), and the laggards (who actively resist buying new stuff).
If you have cool technology and a vision to back it up, he says,
then it's easy to get your first few customers. Then comes the
great chasm: to sell to the early adopters, you need a totally
focused, totally disciplined strategy. That strategy is to pick
a niche market that's big enough to provide a year's revenue,
connected enough to provide a beachhead into much bigger markets,
but small enough that you can concentrate a lot of energy on
providing just what they need, in the total tailored package
that they need it. This explains a lot, and I was impressed
with the analytical rigor with which Moore explains it -- little
wonder that the book has been so influential in the industry.
Even though these books encode the self-interested viewpoint of
a consulting firm, I find it useful to see what follows if we
treat them as descriptions of reality. I am struck, for example,
at the gate-keeping function of the innovators -- those gadget-
oriented people who like the technology for its own sake and
have the resources to buy one of everything they find cool. What
if the world badly needs various gadgets that are deeply useful
but not deeply cool? I am also struck at the uselessness of
classical economics for analyzing the matters that Moore writes
about. He is heavily focused on the construction of meaning
around these systems. His market analysis is psychographic
all the way: segmenting people based on how they think, and
setting strategy in a way that provides each segment with the
facts and stories they need to make decisions according to their
own particular style. The point of selling to innovators, for
example, is not just that they provide easy money, but that they
provide useful technical feedback to the company and valuable
referrals to the early adopters who respect their technical
sense. Likewise, much of the purpose of a niche strategy in
approaching the early majority is that you need to be selling to
people who talk to one another so that you can create a clearly
defined, sustained "buzz" that positions your product to the
exclusion of others. In the end, of course, all books like this
one, however compelling, are only single-factor theories. Heaven
knows how each of these crucial single factors interacts with
the crucial single factors described in all of the other books.
Everything sure seems clear-cut and invigorating from that kind
of distance these books provide, but let's keep in mind that
the day-to-day reality of business is much closer to "Dilbert".
Since I ragged on The Red Herring in TNO 2(12), it has gotten
somewhat better. Its newsstand price has come down by a third
and its copy, now written in large part by venture capitalists,
is less self-serving, or at least less obviously self-serving,
than it had been. Meanwhile, a friend pointed out that the
magazine's title is not completely meaningless; Silicon Valley
types refer to a preliminary prospectus for a start-up company
as a "red herring", after a section of legalese that must appear
in red type to scare away the naive. Still, I really do not
understand the enthusiasm that TRH stirs up among some in Silicon
Valley. It's as if the computer industry has fully internalized
the culture and language of hype, so that the very idea of real
professional journalism cannot even be expressed. Very strange.
In response to my wish for a universal event calendar in TNO
3(4), Steven Hodas <email@example.com> pointed me at Metrobeat,
which covers New York City. As he observes, it's pretty great.
To me, though, it also points to the necessity of critical mass
in such things -- even assuming I lived in New York, I doubt if
I would really make Metrobeat a routine part of my life unless
it included many more event announcements. Its entertainment
coverage is good, though, and you should check it out at
My article on libraries in TNO 3(4) elicited two primary types
of objections. The first, probably best articulated by Larry
Etkin <firstname.lastname@example.org>, is that my call for "information"
tools differentiated by community and genre may cause unnecessary
incompatibility and confusion. Having learned the tools (e.g.,
catalog systems) for one genre of materials would no longer, as
now, enable someone to use the tools for other genres equally
well. The result would be that insiders to a given field have
an even greater advantage over outsiders than they do already.
This is, of course, a legitimate concern. We should recognize
how far things have already gone in this direction, given the
proliferation of proprietary article indexes which must often be
searched independently of one another with different interfaces.
In this way and others, greater heterogeneity of library tools
is inevitable. In response, I think it's important for everyone
with a stake in the matter to codify and encourage good interface
design practices. Perhaps more importantly, we need interface
guidelines like those that came with the original Macintosh, so
that users can carry over as many expectations as possible from
one interface to the next. I use the University of California's
online catalogs quite heavily, but that does me little good when
I am in another library. This situation might well get worse.
Unfortunately it takes a lot of politics, power, and/or luck to
get standards adopted across a whole industry, particularly when
many of the players have already made investments in incompatible
interfaces. I know that many librarians are aware of this whole
problem, and I wish them luck in developing the concepts and
alliances they will need to fix it.
The second common objection to my article, well articulated by
Paul Doty <email@example.com>, is that (contrary to what
I say) libraries do not treat information as a homogeneous stuff.
After all, they already deal with diverse patrons with diverse
needs for diverse materials. Catalogs, moreover, reflect the
insights of cataloguers into the keywords and conceptual systems
of the communities that are interested in particular kinds of
materials. That's all true and important. I just think we
can go further and do better. Catalogs currently make explicit
very little of the internal structure of documents -- not least
because internal structure varies by genre. They also do not
make explicit (though, of course, they reflect implicitly)
the institutional structures that the documents come from and
participate in. Library information is still poorly integrated
with the information maintained by professional societies,
individual research groups, publishers, volunteers, and so
on. As more categories of documents move online; and as more
categories of representations *of* documents move online;
as collaborative work online becomes easier; and as network,
database, and data-interchange standards make it easier to
interlink online representations of documents, it will be
possible for libraries to collaborate more intensively with
one another and with the communities whose work they support.
This will be good.
The Truth and Reconciliation Commission in South Africa has its
home page at http://www.truth.org.za
Jamie Boyle from American University has written a comprehensive
refutation, signed by scores of law professors, of the bogus
legal claims that the Dept of Commerce has circulated in support
of its famous "White Paper" position on electronic copyrights.
David Rothman <firstname.lastname@example.org> has put this material on the
Web at http://www.clark.net/pub/rothman/boyle.htm
The Computer Underground Digest is a just-about-weekly digest
of current information on Internet-based political activism.
Subscription information and archives of past issues can be found
on the Web at http://www.soci.niu.edu/~cudigest/
Paul Edwards <email@example.com> has gathered a set of
well-documented Internet history material on the Web. Here is
(uses Netscape 2.0 frames)
Paul's valuable institutional history of artificial intelligence,
"The Closed World: Computers and the Politics of Discourse in
Cold War America", has just appeared from MIT Press.
The US Department of Education has published a series of fairly
useful White Papers on "The Future of Networking Technologies
for Learning" at http://www.ed.gov/Technology/Futures/toc.html
Phil Agre, editor firstname.lastname@example.org
Department of Information Studies
University of California, Los Angeles +1 (310) 825-7154
Los Angeles, California 90095-1520 FAX 206-4460
Copyright 1996 by the editor. You may forward this issue of The
Network Observer electronically to anyone for any non-commercial
purpose. Comments and suggestions are always appreciated.
Go back to the top of the file