nettime: The end of information and the future of libraries - Phil Agre

[Pit Schultz <pit@contrib.de>]

Date: Thu, 9 May 1996 22:06:18 -0700 (PDT)
From: Phil Agre <pagre@weber.ucsd.edu>
Subject: TNO 3(4), worth the wait.
X-URL: http://communication.ucsd.edu/pagre/rre.html

[...]

The end of information and the future of libraries.

  My work is thinking about basic ideas of technology in ways that
  let us see them as products of social processes, and as part of
  social processes.  For example, computing has very particular
  ideas about how to represent human activities.  These ideas have
  histories.  They could be different, and they have significant
  consequences for privacy.

  Let us consider another basic idea of computing, information.
  We all think we know what information is.  Computer people and
  librarians both define their work in relation to something they
  call information.  But I want to suggest that information might
  be an obsolete concept, and that emerging technologies are
  yelling in our ears to move along to other, different ideas.

  What is information?  We can define it in a narrow technical
  way.  Shannon defined one notion of information in his theory
  of the capacity of a communications channel; information for
  him is measured in bits, and each bit is a distinction that
  is meaningful to the parties on each end of the channel.
  Bateson said something similar when he defined information
  as differences that make a difference.  Computer people
  often speak of information in terms of the states of digital
  circuits that represent binary states of affairs in the world.

  In each case, information is an idea that builds a bridge
  between the states of artifacts and meanings in people's lives.
  We often hear that this is an information age, or an information
  revolution, or that information rather than capital is now
  driving the global economy.  It is not at all clear what any of
  this means.  I think that in practice we tell three stories to
  ourselves about information.  Each story profoundly affects our
  thinking by encoding particular views us about the relationship
  between designers, information users, and information itself.  I
  will refer to these stories as information processing, masculine
  transcendentalism, and information professionalism.

  (1) Information processing

  Computers originate in automation; "computer" was originally
  a job title, not a machine.  Early computing methodologies
  were modeled on industrial automation methods -- a flowchart
  is really an industrial process chart.  When you hear the phrase
  information processing, therefore, I want you also to hear
  phrases like food processing and sand and gravel processing.
  Information, according to this story, is an industrial material
  like corn or oil or metal.

  The information processing story assigns particular roles to
  designers, users, and information:

   designers - gods
   users - factory machines
   information - processed material

  (2) masculine transcendentalism

  I take this marvelous phrase, masculine transcendentalism, from
  the historian of technology David Noble.  We can see masculine
  transcendentalism at work in Wired magazine, or in all of the
  hype around artificial intelligence or virtual reality.  The
  story is this: someday soon, the physical world is going to
  wither away.  Everything is going to become digital.  All of our
  minds will be downloaded onto machines.  All of our books and
  paintings will move into digital media.  We will no longer have
  bodies, and most amazingly of all, we will work in the paperless
  office.  Noble's brilliant insight is that this is a religious
  worldview, and his historical research demonstrates compellingly
  that it developed *out of* a religious worldview without any
  particular discontinuity along the way.  It is a millenarian
  worldview in that it posits a perfect future in which everything
  will be transformed.  It is a transcendental worldview in that
  it calls for the whole world to be raised up and dissolved into
  incorporeal realm that leaves the body and all the messy stuff
  in the social world behind.  It sounds funny and hyperbolic when
  you frame it this way, but it is an enormously influential way
  of speaking in industry and elsewhere.

  Here, then, are the basic relationships posited by masculine
  transcendentalism:

   designers - prophets
   users - caught up in an inevitable rapture
   information - the fabric of heaven

  (3) information professionalism

  Information professionalism is a story that both computer people
  and librarians tell, but I want to focus on the librarians'
  version here.  This story goes: we are professionals; there is
  this stuff called information; and our professional expertise
  consists of managing large bodies of information and connecting
  people with information.  These professionals are generalists,
  or specialized at most to very broad areas, and libraries treat
  very disparate kinds of stuff in the same way.  This view is
  understandable when you have a dozen librarians in a library
  building, and they are buying, cataloguing, and managing
  information that a hundred different kinds of people are using.
  The librarians need to routinize their work, and they need highly
  rationalized, detailed procedures so that the product of their
  work -- a catalog, for example -- is uniform and so that this
  product can be produced efficiently.  Libraries have themselves
  been factories in many ways -- thousands of books just have
  to get catalogued.  None of this is a criticism of librarians,
  who have been working within the constraints of particular
  technologies and institutions.

  Here, then, are the relationships that the information
  professionalism story posits:

   designers - professionals
   users - individuals with information needs
   information - homogenous stuff to be stored and retrieved

  I do believe that information technology is contributing to a
  major change in the world, but I think that this is precisely
  a change that makes each of these stories obsolete.  The old-
  fashioned factory story is already under heavy attack -- we've
  automated an awful lot of tasks already, and the resulting
  machinery requires a lot of skill and expertise to use.  But it
  is striking that we haven't often questioned this view in the
  context of information.

  Masculine transcendentalism, for its part, is really one of those
  yesterday's tomorrows, like the Jetsons.  If we look at what is
  really happening in the world, we see information technology as
  a nervous system for the physical world, not as a replacement for
  it.  (See, for example, TNO 1(5).)

  But it's information professionalism that I really want to
  focus on.  The problem with information professionalism is
  really a problem that the others share underneath: it treats
  information as a homogenous substance.  A good way to think
  about information is that it's the professional object of
  librarianship.  Every profession has its object: for law
  everything is a case, for medicine everything is a disease,
  and for librarianship everything is information.  In each case,
  someone walks in the door with a problem, and the professional's
  job is to find their object in that problem, and to talk about
  the problem in a way that makes it sound like a case, a disease,
  or information that can be compared with other cases, other
  diseases, or other information.

  There's a deep trade-off: each profession achieves generality by
  reducing everything to a common denominator, leveling everything
  to common terms.  Each profession can help everyone, but they
  cannot help them very well.  Library materials are indexed in a
  very sophisticated way -- certainly much more sophisticated than
  the keyword searches that prevail on the Internet -- but it is
  one uniform indexing scheme, despite the many different places
  that different patrons might be coming from in their lives.

  We can think about solving this problem by using information
  technology to support several different coding schemes, and I
  think this is a good thing to do.  But I want to back up and
  suggest a more radical approach.  Let's get beyond the stories
  we have told ourselves about information and tell different
  stories about different sorts of objects.

  I want to suggest that the defining feature of our new world
  is that people talk to each other, a lot, routinely, across
  distances, by several media.  It makes no sense any more to
  ask how individuals use information.  Instead, let us ask how
  communities conduct their collective cognition.  Let's define
  a community, as per TNO 2(7), as a set of people who occupy
  analogous structural locations in society.  The residents
  of Palo Alto are a community, but so are cancer patients,
  corporate librarians, and people who are in the market to buy
  any particular sort of product.  Emerging technologies allow
  communities to think together.  The fact that cancer patients
  can think together is already turning medicine inside-out.  The
  fact that customers in computer-related markets talk intensively
  to one another on the Internet is increasing the amount and
  variety of information in the marketplace.  The future, in my
  view, belongs not to information but to this active process of
  collective cognition in communities.

  It might be objected that we will always have libraries
  and bookstores, and they will still be full of information.
  But that's not the best way to look at it.  The first thing
  that library cataloguing schemes lose is the dialogic nature
  of articles and books: they are all turns in a conversation,
  responding to a particular literature or cultural background
  and addressed to a particular audience.  Every community conducts
  its collective cognition through diverse mechanisms, from rumors
  to conferences to newsletters to wandering bards to Internet
  mailing lists to articles and books.  The library is one window
  on this whole dynamic interplay, but it is not a window that
  lets us see that dynamic interplay very clearly.  Perhaps it is
  an artificial window, a means to serve a subset of "information
  needs" that is largely an accident of past technologies and
  institutions.  Many different kinds of energy pass through
  the library, but the library reduces them all to information
  retrieval, a homogenous category that it can work with.

  The solution, I think, is not to pave the cowpaths by automating
  the institutions we have now.  Instead, I think we should explore
  the full range of means by which we can support the collective
  cognition of communities.  Every community has its own mix of
  communications mechanisms, its own history and institutions, its
  own symbols and vocabularies, its own typified activities, its
  own constellation of relationships, and perhaps most importantly,
  as TNO 2(11) suggests, its own genres of communicative materials.
  If we want a focal concept to replace information, we might
  want to choose genres.  Genres are stable, expectable forms of
  communication that are well-fitted to certain roles in the life
  of some particular communities.  Business memos, opinion columns,
  action-adventure movies, Interstate Highway signs, business
  cards, and talking-head TV political shows all have stable forms
  that evolve to serve needs in the midst of particular activities.

  I don't think we should be automating information professionals
  out of business.  Quite the contrary, I think we should be
  giving them a bigger job: reaching out to support the collective
  cognition of particular communities.  This might include systems
  to support the creation, circulation, and transformation of
  particular genres of materials.  It might include setting up and
  configuring mailing lists or other, more sophisticated tools for
  shared thinking.  It might include both face-to-face and remote
  assistance.  Distributed alliances of librarians might support
  specific distributed communities, while comparing notes with one
  another and sharing tools.

  This view has many consequences.  It follows, for example,
  that a digital library isn't one big system but a federation of
  potentially quite different systems, each embracing a range of
  functionalities and fitting into people's lives in potentially
  quite different ways.

  It also follows that each community will have, to some extent,
  its own infrastructure with its own evolution.  Standards are
  crucial.  Tools for shared thinking work best when everyone is
  using them, and so supporting a community's transition to new
  tools will require consensus-building, well-timed coordination,
  training, and a shifting division of labor between professional
  librarians -- or, as we might start calling them, communitarians
  -- and mutual aid and self-help among a community's members.  No
  more factories, no more millenarian fantasies, no more isolated
  information warehouses.  Instead, perhaps, we might be able
  to build, and help other people to build, the interconnected
  pluralistic society that we so badly need.

[...]

# Pit Schultz, Kleine Hamburger Str. 15, 10117 Berlin, pit@contrib.de


--
*  distributed via nettime-l : no commercial use without permission
*  <nettime> is a closed moderated mailinglist for net criticism,
*  collaborative text filtering and cultural politics of the nets
*  more info: majordomo@is.in-berlin.de and "info nettime" in the msg body
*  URL: http://www.desk.nl/nettime/  contact: nettime-owner@is.in-berlin.de