The Information Railroad Is Coming
Wm. A. Wulf
University of Virginia
It's New Year's Day, 1895. My name is Hans. For seven generations
my family has made the finest buttons in the region, using the good local
Today I learned that the railroad is coming to our village. My friend Olaf
says that cheap factory buttons will come on the trains, but they will
never compete with my craftsmanship.
I think he is right, and wrong. They will come, but they will compete with
my buttons. I must make some choices. I can become a distributor for the
new buttons, or I can invest in the machinery to make buttons and export
them. Or, closest to my heart, I can refine my craft and sell exceptional
buttons to the wealthy.
My family's business is dead. I cannot stop the train; I must change.
Universities are in the information business, and the information railroad
With 20-20 hindsight it's easy to accept the demise of a quaint industry
-- or more accurately, the demise of a quaint method of manufacture, distribution
and sale. The button industry flourishes, of course. Even the craft of hand
made buttons is doing well, if my local art fair is any indication. However,
the nature of the industry changed dramatically as technology allowed the
manufacture and distribution of vastly less expensive but highly serviceable
It's harder to look inward at the university, with its tradition and obvious
social value, and introspect about whether it might change in dramatic
ways. But, although their roots are a millennia old, the university has
changed before. In the early nineteenth century it embraced the notion
of secular, "liberal" education. In the late nineteenth century
it included scholarship as an integral part of its mission. After World
War II it accepted an implied responsibility for national security, economic
prosperity, and public health in return for federally funded research. Although
the effect of these changes have been assimilated and now seem "natural",
at the time they involved profound reassessment of the mission and structure
of the university as an institution.
Forces are always acting on universities. Some of them, notably the political
ones, have great immediacy and hence get a good deal of attention . In
the current milieu these include the reassessment of the rationale for federal
funding of research, the desire for greater "productivity" from
the Professoriate, and so on. In spite of the attention accorded these issues,
I believe information technology has a far greater potential to provoke
fundamental change in our system of higher education. Moreover, I am certain
these changes are much closer than most people realize.
Let me be clear. Higher education will flourish, just as does the button
industry. If anything, the need for advanced education is increasing simultaneously
in multiple dimensions. A greater percentage of the world's population needs
to be educated to be productive in an increasingly technological workplace.
The period during which particular skills are relevant is shortening, and
so the need for "lifelong learning" is increasing. The knowledge
and skills necessary to function at the frontier of knowledge are increasing
as well, and so with it the need for advanced degrees.
Higher education is not in danger. But we would be wise to ask whether the
particularly quaint way that we manufacture, distribute and deliver that
education will survive the arrival of the information railroad. They may,
but I don't think so. I think there will be major changes - changes not
only in the execution of the mission of universities, but in our perception
of the mission itself.
As Clark Kerr pointed out , the relative ranking of universities changes
slowly, but there are times that are more propitious than others for change,
and the next decade is one such. Because of the speed with which information
technology is advancing, decisions are being made now - or more likely,
defaulted now - that will have material effect on the real and perceived
quality of institutions of higher education. In my experience almost none
of the current generation of senior university administrators understand
what is happening.
Thus, I feel we must engage in an intellectually honest exercise of understanding
the implications of technology on our institutions - at least as best we
can. Stimulating that discussion is my purpose here. But, before proceeding
we need to dispense with two issues: First, is the button analogy valid?
Second, is the technology for higher education really going to change all
Certainly some specifics of the button parable are inappropriate - universities
don't make a product that can be mechanically mass produced, for example.
But while pointing out the differences, it would be a mistake to dismiss
the similarities. Both are highly labor intensive and depend on the skill
of their master craftsmen. Both have been regional, requiring collocation
of the producer and customer. Both have long traditions. Both contributed
to the prestige of their locale. Both evolved powerful guilds to protect
the masters. And, now the university is also faced with a technological
Universities share at least some of the attributes of other vertically-integrated
industries as well. We "manufacture" information (scholarship)
and occasionally "reprocess" it into knowledge or even wisdom,
we warehouse it (libraries), we distribute it (articles and books), and
we retail it (classroom teaching). Information technology has already changed
each of these, and the future change will be much greater. Like industries
that have been overtaken by technology, we need to understand its individual
and collective impact on our basic functions. It's not a comfortable thought,
but we must at least consider that a change in technology - a change that
will facilitate the flow of our essential commodity, information - might
provoke a change in the nature of the enterprise.
As for information technology - one of the hardest things for most people
to understand is the effect of its exponential rate of improvement. For
the last four decades the speed and storage capacity of computers have doubled
every 18-24 months; the cost, size and power consumption have become smaller
at about the same rate. The bandwidth of computer networks has increased
a thousand-fold in just the last decade, and the traffic on the network
continues to grow at 300-500% annually. For the foreseeable future, all
of these trends will continue; the basic technology to support them exists
The compound effect of this rate of improvement is hard to appreciate -
but, speaking of ENIAC, the first fully electronic digital computer, a
1949 article in Popular Mechanics said:
" ENIAC contains about 18,000 vacuum tubes and weighs 30 tons, but
in the future computers will contain only about 1,000 tubes and weigh
only 11/2 tons."
Thirty five years later a typical microprocessor is about 100,000 times
more powerful, contains the equivalent of ten million tubes and weighs substantially
less than an ounce. Imagine the nanoprocessor of the a few decades hence.
To my knowledge, there has never been a similarly rapid, sustained change
in technology, especially one with such broad social application. By comparison,
even the industrial revolution seems modest in scope and leisurely in pace.
Lacking a precedent, we need to work harder to imagine the impact of future
computers and networks. Thinking about the current ones, in fact, can be
misleading; it's all too easy to assume that something won't change just
because today's technology doesn't support that change. Instead, it's almost
better to hypothesize a change and then ask how soon the technology will
support it; the answer will often be surprisingly soon.
Don't think about today's teleconferencing technology, but one whose fidelity
is photographic and possibly three dimensional. Don't think about the awkward
way we access information on the network, but one in which the entire world's
library is as accessible as my desktop files. Don't think about the clumsy
interface to computers, but one that literally listens and talks in your
jargon, not mine. Don't think about the storage on today's PC, but one
with terabytes (millions of megabytes). We can't afford it now, of course,
but that is the power of the equipment that will be affordable in a decade
or so. That is the equipment that will shape the future of the university.
How will we use this equipment to change education and scholarship? That
seems like a simple question, but as both an academic and a Computer Scientist,
I don't know. Certainly knowledge, its creation, storage and communication,
is part of the essence of a university. The ability to process information,
the "raw stuff" of knowledge, thus sits at the heart of the university
mission. A technology that will alter that ability by orders of magnitude
cannot avoid having a impact on at least how we fulfill our mission - and
possibly on the mission itself.
To borrow a phrase from the business world, information technology is a
"core competence" of our industry; the leaders in developing and
exploiting it will be the leaders of the next century. This is one of Kerr's
Perhaps as a start we might look at several functions of our vertically
integrated information business and note how they have been and might be
Scholarship: The impact of information technology on science is both
apparent and pervasive. Scientists now routinely talk of computation as
the "third modality" of scientific investigation, on a par with
theory and experimentation.
The easy examples are those that simply automate what was done manually:
the reduction of data, the control of instruments, etc. The profound applications,
however, are those that lead to whole new areas of research and new methods
of investigation - and thus to science that was not, and could not be done
before: the final proof of the four color conjecture, analysis of molecules
that have not been synthesized, measuring the properties of a single neuron
by growing it on a silicon chip, watching a model of galaxies collide, and
letting a scientist "feel" the forces as a drug docks in a protein.
These applications have transformed the nature of scientific investigation;
they led to questions that would not even have been asked before.
I don't think science, however, will be where we see the most dramatic impact.
I say that despite a recent report from the National Research Council
that I helped coauthor - a report that paints an expansive image of the
transformation of scientific research. Instead, I believe that a more dramatic
transformation is about to shake the foundations of scholarship in the liberal
arts. Humanists more than scientists will lead the way to innovative applications
of the technology in the university.
The comfortable stereotype of humanists as technophobic just doesn't apply
anymore. The availability of both text and images in electronic form coupled
with the processing power of modern computers allow the humanist to explore
hypotheses and visualize relations that were previously lost in the mass
of information sources. The presentation of humanists' scholarly results
in electronic form is moving even faster. Precisely because of the complexities
of the relationships they need to present, the subtle webs of relation and
inference they need to express, electronic "hypertext" books and
journals are emerging. Indeed, they are emerging faster, with more vigor,
and with more effect on their disciplines than their counterparts in the
We all expect scientists and engineers to use computers in their research,
but the notion that information technology could be central to humanistic
scholarship is a bit more startling - at least to me. It was in large measure
talking about the application of computers to historiography and the theory
of text that opened my eyes to the larger issues that I am trying to raise
Textbooks: I don't know anyone who prefers to read from a computer
screen, and besides you can't take a computer to the beach - or so say the
nostalgic. They are right, and yet so profoundly wrong.
There are two fallacies here. The first is the assumption that electronic
books will contain only text, and hence be the essentially the same as paper
books but presented differently. In reality, it will not be possible to
reproduce an electronic book on paper. They will not be a simple linear
presentation of static information. They will contain animation and sound.
They will let you "see the data" behind a graph by clicking on
it. They will contain multidimensional links so that you can navigate through
the information in ways that suit your purpose rather than the author's.
They won't contain references to sources, but the source material itself
- the critique of a play will "contain" its script and performance.
They will have tools that let you manipulate the equations, trying them
on your own data or modifying them to test scientific hypotheses. They will
let one annotate and augment the documents for use by later readers, so
making it a "living document".
The second fallacy is presuming today's technology. We should not be talking
about reading these electronic books from today's screen. The advantages
of the electronic book will be so strong that engineers will make the "form
factor" of the medium humane. Screens already exist in the laboratory
with a resolution about the same as the paper you are reading right now
- as do flexible ones of somewhat lower resolution. Why would anyone lug
around several heavy books when something the size, clarity and weight of
a single one contains them all? I mean them all - all the ones in
the Library of Congress. I will take my computer to the beach!
Libraries: For thousands of years the focus of libraries has been
on the containers of information, books. The information itself was
the domain of the library's users, not the library. Information technology
turns that premise on its head, and with it many of the deepest unstated
assumptions about the function of a library.
Tracing back to Alexandria and before, the principal objective of librarians
has been to build the collection - to amass a set of materials was their
measure of worth. But, in the future a library will not "collect".
Electronic information can be communicated virtually instantaneously, so
its source location is irrelevant. Instead of a hoarder of containers, the
library must either become the facilitator of retrieval and dissemination,
or be relegated to the role of a museum.
If we project far enough into the future, it's not clear whether there is
a distinction between the library and the book. The "technology"
of the bibliographic citation pales by comparison to the hypertextual link
- to the ability to gain immediate access to the full referenced source,
and hence to browse through the context of the reference. It will take a
long time to build the web, and especially to incorporate the paper legacy,
but the value of a seamless mesh will doom the discrete, isolated volume.
As the library and the book merge, it seems compelling to me that another
merger will accompany it - a merger precipitated by devolving disciplinary
boundaries. Knowledge isn't inherently compartmentalized; there is only
one nature, there is only one human record. The division of the sciences
into Physics, Chemistry, etc., and their further subdivision into Physical
Chemistry is a human imposition, as is the division into History, English
and Anthropology. For very practical reasons, paper texts have mirrored
this artificial division, but those "practical" reasons evaporate
in the electronic world. Clearly the "long pole in the tent"
will be human rather than technical; disciplines are complex and idiosyncratic
social structures that will not easily dissolve. However - and here I can
only speak with the even the smallest authority about technological disciplines
- much of the most interesting work is already happening at the boundary
of traditional disciplines. That's not new news; Einstein opined that most
of the important science lay at the interstices of traditional disciplines.
What is new is that we have a technology that facilitates incremental accretion
of knowledge at these interstices.
Finally, note that the book as we know it is passive; they sit on shelves
waiting for us to read and interpret them. While there is an intellectual
thrill in discovery and interpretation, passivity of the text is not required
for that. As Marvin Minsky, a Professor at MIT, said: "Can you imagine
that they used to have libraries where the books didn't talk to each
other?". One of the profound changes in store for libraries is that
parts of their collection will be active, software agents collecting,
organizing, relating and summarizing on behalf of their human authors. They
will "spontaneously" become deeper, richer, and more useful.
Teaching: The notion of computer-aided instruction has been touted
for thirty years. Frankly, it has had relatively little impact, especially
at the university level. The reason is obvious: chalk and overhead projectors
have been perfectly adequate technology given the current nature of scholarship
If, however, the bulk of the Professoriate are using information technology
in their scholarship, and the results of that scholarship can only be exhibited
using the technology, the classroom will follow rapidly. How will it follow?
Not, I think, by the "automated drill" scenario we have come to
associate with Computer Aided Instruction, CAI.
Beyond automated drill, the obvious application of technology is telepresence
- the possibility of involving remotely sited individuals in a seminar,
for example. Again, do not think in terms of today's teleconferencing technology;
as the fidelity of communication improves, telepresence will certainly
happen. While now it's a big deal to bring a leading authority to campus,
and access to the person is often limited to research colleagues and graduate
students, this will not be the case in the future. The technology will give
an increased number of undergraduates access to these authorities. Removed
from the overhead of travel, who among us would not cherish a few hours
each week with the bright young minds at a remote Harvard or Yale?
These are interesting but mundane applications - mundane in the sense that
they do not change the educational process in a deep way.
More fundamental is the opportunity to involve students in the process
of scholarship rather than merely its results. We like to say that we "teach
students to think, not merely to learn rote facts", but in truth
- we mostly limit them to thinking about what has been thought before. We
can't ask them to explore new hypotheses, because of the practicalities
of access to sources and the sheer grunt work of collecting and analyzing
data. Information technology eliminates those "practicalities".
- they are forced through the linear sequence of the text, course and curriculum
before we judge that they "know enough" (facts) to embark on a
scholarly project (think).
A hint of this kind of change can be detected in a report in the Chronicle
of Higher Education about the impact of the release of Thesaurus Linguæ
Gaæce on scholarship and education in the classics. The article
noted that the release of this database, which [now] includes virtually
all Greek Literature from Homer through the fall of Byzantium, has enabled
undergraduate participation in research.
One cannot leave the subject of teaching without at least mention the subject
of "productivity" - the current code word to capture the public's
frustration with the rising cost of college education and the perceived
emphasis of research over teaching. Its simplistic translation is to have
professors spend more time in the classroom and less in the laboratory.
Particularly given the recent wrenching restructuring of industry, the
public has ample cause to ask why an elitist academe should be exempt from
a reorientation toward greater customer satisfaction.
The irony, of course, is that one of the oldest figures of merit for any
school - a low student/teacher ratio - is diametrically opposed to "productivity".
Information technology is not going to resolve this tension; in the end,
for our own children we want relatively individual attention from the most
qualified, intellectually alive professoriate possible. Information technology
can, however, shift the focus of the discussion to the effectiveness
and quality of the student/teacher interaction rather than just the number
of contact hours.
Indeed, in modest ways it already has shifted that focus. By removing the
barriers of both space and time for example, email has given my students
have much greater access to me than ever before. Involving students in the
process of scholarship and giving them greater access to international authorities
are more profound shifts, but I suspect that these are still just pale precursors
of what we can do. Part and parcel of rethinking the impact of technology
on the university is addressing precisely this issue.
Let's return to the main thread. Whether or not you agree, I hope that
this discussion has at least suggested that activities in the academy might
change. Even so, that does not imply that the nature of the university
as a whole will change. Will it?
One approach to such a question is to examine unstated assumptions; that's
hard, but I would like to examine just one.
Historically a university has been a place. The stone walls of St.
Benedict's cloister at Monte Cassino were the bastion that provided defense
against the physical and intellectual vandals of the dark ages. In colonial
times, Jefferson's Academical Village provided access to scholarly materials
as well as collegial interaction by collocation. In contemporary times,
scholars flock to scientific instruments and library collections. And, where
the scholars assembled, the students followed.
In his influential nineteenth-century essays on the Idea of a University,
Cardinal John Newman wrote :
If I were asked to describe ... what a University was, I should draw
my answer from its ancient designation of a Stadium Generale ... This description
implies the assemblage of strangers from all parts in one spot.
The Cardinal then goes on at some length to emphasize that books were an
inadequate source of true education that must be buttressed with discourse
- which is obviously only feasible if the discussants are collocated. Thus
the notion of being "in one spot" is, to him, essential
to the very definition of the university; as he says, "... else,
how can there be any school at all?"
But, with the possible exception of teaching, to which I'll return in a
moment, I believe that information technology obviates the need for the
university to be a place.
Once again, please remember that although we are presuming technology better
and cheaper than today's, it is not hypothetical- the trends are clear,
the capabilities for at least the next decade are predictable, and in many
cases the technology is already in the laboratory. Only how we will use
the technology is at question.
With this powerful, ubiquitous computing and networking, I believe that
each of the university's functions can be distributed in space, and possibly
in time. Remote scholarship and authoring are the direct analogs of telecommuting
in the business world, and every bit as appealing. Academics tend to identify
more closely with their disciplinary and intellectual colleagues than with
their university. Freed from the need to be physically present in classroom,
laboratory or library, grouping by intellectual affinity may be more appealing.
But even then, physical grouping may be unnecessary and even undesirable
as such things as location preference are taken into account.
There are some disciplines that need shared physical facilities, say a telescope,
that suggest the need of a "place". But note that large scientific
instruments such as telescopes and accelerators are already run by consortia
and shared by the faculty from many universities, and many of these facilities
do not require the physical presence of the investigator - they could be
"on line" and accessible via the network. Indeed some instruments,
such as those for space physics at Sondre Stomfjord in Greenland, are already
accessed on the Internet. The university as "place" is already
irrelevant to at least some scientific scholarship.
As with instruments in the sciences, direct access to archival materials
is necessary for some humanistic scholarship - but hardly all, and certainly
not all of the time. Ponder the excitement, for example, caused by the
recent release of the images of the Dead Sea Scrolls even though the scrolls
themselves are not accessible to most scholars. If anything, the ubiquitous
information infrastructure will provide greater access to archival materials
to a much larger set of scholars, of a quality that's "good enough"
for most purposes.
As for teaching, we don't really know whether it can be distributed or not.
I do know that even asking the question is considered heretical by some
good teachers - teachers who contend that physical eyeball-to-eyeball contact
is necessary. Others, including me, contend although they need feedback
to teach well, there is a threshold of fidelity beyond which one does not
need to go; student and teacher probably don't need to smell each other,
for example. Thus, there is some finite amount of information required
to produce an adequate representation of the parties. If true, when that
threshold of fidelity is reached electronically, high quality teaching will
be distributed. The fallacy in Cardinal Newman's reasoning was only that
he could not imagine quality discourse at a distance - but that is precisely
what the technology will enable.
Can an institution such as universities that have existed for millennia
- icons of our social fabric - disappear in a few decades because of technology?
Of course; if you doubt it, check on the state of the family farm. Will
the "university as place" in particular disappear? I expect not;
the reduced importance of place does not imply no place. However,
just as farming has been transformed, so will the university. The everyday
life of both faculty and students will be very different.
I have more questions than answers as to the new shape of the new university.
Having now laid the groundwork, let me pose a few of them:
- I believe that higher education will not only survive, it
will flourish. But, are the choices for universities, like for Hans, to
become mass market manufactures or distributors, or niche tutors to the
- Does it really make sense for every university to support the full complement
of disciplines, or should they specialize and share courses in cyberspace?
This might be a natural consequence of aggregation by disciplinary affinity,
- Might professors affiliate with several institutions, or become "free
lance" tutors to telepresent students? Indeed, might we return to
the future of tele-itinerant scholar/tutors?
- Might some employers (and hence students) prefer a transcript that lists
with whom certain courses have been taken rather than where? Presumably
a student that has taken a computer architecture course from Patterson (at
Berkeley) and a compiler course from Kennedy (at Rice) would at least appear
to be better trained than one that took all their courses at just one of
these fine schools.
- What about alumni and sports? Surely the allegiance of alumni to their
alma mater has a great deal to do with place and is cemented on Fall
weekends; since the support of alumni is essential to universities, isn't
that very human need sufficient to perpetuate university as place? Perhaps.
But broad alumni support has become essential to the university only in
relatively recent times. Moreover, alumni associations and large sports
programs were created to support the university as place, not the other
- Will universities merge into larger units as the corporate world has done,
or will the opposite happen? I can argue either side of this question. On
the one hand, if a university isn't (just) a place, its major remaining
function is certification - it certifies the competence of the faculty,
programs and graduates. We don't need thousands of organizations to do that.
On the other hand, I can envision many small colleges being empowered to
provide a broad curriculum via telelocation while retaining the intimacy
so valued in our small liberal arts institutions. I don't know anyone that
really wants the impersonal ambiance of a mega-university. The current
size of these universities seems optimized for the physical infrastructure,
not either education or scholarship.
- Might the technology revive the talented amateur's participation in the
scientific community? Except for a few disciplines like astronomy, the talented
amateur has largely disappeared from scholarly discourse in science and
engineering. Surely such individuals still exist, but they are isolated
from the community of scholars. How can/should the university re-engage
- What about the various businesses that have affiliated with universities
- the university press being an especially poignant example? My guess is
that each of these will be forced to rethink its principal mission, and
many will be irrelevant.
- Will more (most?) universities serve a global clientele, and how does
that square with the publicly supported university in the US? In particular,
will private universities have greater flexibility to adapt to globalization,
thus dooming the public universities?
- Does the function of socializing young adults, which perhaps remains a
reason for "place", need to be coupled with the educational function,
or could it be done better by some form of social service?
Some will interpret these questions as threatening; I don't. That there
will be a change seems inevitable. But, change always implies opportunity,
and in this case the opportunity to improve all facets of what we do in
the academy. The challenge is to anticipate and exploit the changes.
Universities are in the information business, and the information railroad