The images you see now are among the oldest known artistic expressions of our species. They are the cave paintings from Altamira, in the province of Cantabria, Spain, estimated to have been painted about 18,000 years ago. The artist made use of natural protrusions of the cave roof to give depth to the images. He or she made pigments from charcoal, from clay, and from stones, employing various techniques to apply the colors. From these we infer artistic intent. These pictures of buffalo, horses, and elk are evidence of our primordial ancestors coming to grip with their world. There are other drawings in the cave that do not speak for themselves. What do these “tectiforms” represent? Can we hypothesize that these humans spoke, that they counted, that they communicated? Of course we can, but there are only images, these human expressions, for proof.
On the other screen are some simulations and some images of planetary surfaces depicting the Voyager space probe project. The simulations obviously show the spacecraft in perspective with distant planets and moons, but the surface image of the Neptune’s moon, Triton, represents image data sent back from Voyager. These craft are scientific expressions of man’s desire to know, to explore, to describe. Because each craft carries discs of gold inscribed with images of man, woman, and child as well as some of our music and math, they too are examples of human communication. We receive and interpret data sent back to us from space, as the flyby over the surface of Triton shows. And, as we all know, numerous other craft have been sent and more will be sent into space for exploration, each sending back signals of what their mechanical sensors have perceived.
Between these extremes - 18,000 year old cave paintings,
and streams of data sent back to Earth - come the entire range of the
human record of expression, all which are potentially the responsibility
of our societies’ cultural custodians, namely museum curators,
librarians, and archivists. And whether such representations are transmitted
on paper or by bits and bytes, they are in principle ideas worth saving
by one generation, for transmission over time, to subsequent generations.
This is our work, librarians'
work:
the selection and organization of ideas, expressions and knowledge
for distribution from the past, for the education and edification of
those alive now, as well as for people not yet born.
It is good to reflect on this work on the occasion of the 400th anniversary
of the opening of the Bodley
Library, at a time that has seen human communication and expression
greatly perturbed by the rapid growth and adoption of the Internet
as a new channel for expression and the transmission of ideas. This
celebration of the founding of the Bodley Library and the persistence
of the Bodley’s librarians, keepers, and masters at Oxford is
an auspicious time to look
forward as well as back. That is my task today: to look forward,
to speak about the future of the great research libraries. It is a
great honor to be asked to do so, and I thank Reg Carr for extending
the invitation.
As I began to read and reflect on this assignment, I did the usual
review of the literature. Imagine my horror to discover that there
are well over 3,850 hits
on a Google search of the phrase “future of libraries”,
though only four hits
on the phrase “future of great research libraries,” all
referring to this talk at this conference. A quick search of the periodical
literature produced about 500 entries for articles, notes, and so forth
on the more inclusive phrase, few of which duplicated the Google search
results. So, with a little license, one might say that the recent literature
comprises over 4,000 entries.
While I
considered my topic, I pondered what we might mean by the phrase “great
research libraries” and why these merit discussion as having
a future apart. I was mindful the immediate audience would include
colleagues and friends who lead research libraries which they and others,
myself included, might describe as “great.” In a sense,
I think of this as a genuine superlative, that is, not as a comparative
term.
Though there is substantial presumption that size alone makes a great
research library, another measure might be longevity. In my view that
list would start with the Biblioteca Apostolica Vaticana. The U.K.’s Consortium of
University Research Libraries has
a nice set of criteria for membership that
could be applied as well:
What then makes a great research library? Here are my
criteria:
•a very large collection, both in numbers of volumes as well as in number
of disciplines, topics, languages, and literatures
•a distinctive, broad, and extensive collection, with prominent special
collections and substantial annual rates of acquisitions
•highly qualified subject and technical professionals on staff
•a lengthy history of library operations
•currency with changing methods and materials of scholarship by collecting,
organizing for access, and preserving new research resources
•an active program of preservation and conservation
With these criteria in mind, I would guess that we could today construct
a list of about 40 or 50 “great
research libraries” in the world, several of them national
libraries, many university libraries, especially including the Bodleian,
and a very few public libraries. Great
research libraries also have squads of specialist staff, not limited
to specialist librarians, bibliographers, and archivists, but also
including exhibit curators, publishers and editors, security officers,
and … conference organizers. For the purposes of this presentation,
all other libraries fall into a taxonomy no one has patience to develop,
to describe, or to discern with precision.
Harold
Billings, our distinguished colleague who is head of the
Libraries of the University of Texas at
Austin, adds a telling qualification:
“Sustaining a great library does not diminish other libraries. It adds
to, builds up, and enlarges the capabilities of other libraries, and is itself
strengthened in turn through its collaborative association with others.”
To speculate on the future of the great research libraries, one should
consider what research might be done in the future as much as determine,
through our selection of materials for our collections, what research
will be possible. These predictions of what some scholars might wish
to do are useful in shaping the programs of great research libraries
to accommodate and support them.
Here are three sample scenarios based on what scholars are asking for
now and what will be possible soon. Each scenario is set in the second
decade of the 21st century. There is one for the humanities, one for
the social sciences, and one for the sciences. Others could easily
be constructed for other domains.
The first scenario is that of a musicologist investigating
the simultaneous use of jazz idioms and stylistic elements by French
classical musicians
of the first half of the 20th century and the use by American jazz
musicians of French musical elements in the same period. In addition
to the core musical questions - of properly identifying stylistic elements
like harmonic structures, use of specific chords, rhythmic patterns,
formal structures, distinctive decorative elements, modal melodies,
and more or less clear quotation - there are social, economic, and
perhaps even political questions entwined.
In order to pursue this line of research on the core musical question,
the scholar will need to have listened extensively to the easily available
works from the repertories in question. She will then convert exemplars
of the styles to transposable, numeric “signatures” in
digital form. She will then run a program comparing her signature elements
to the thousands of recorded excerpts she will have collected.
Automatic searches with the specifications of composers, performers,
groups, styles, critics, performing locations, and titles of works
will have been devised to search the contents of digitized newspapers
of the period, as well as magazines, histories of music, memoirs and
collections of letters, and archival documents of the key composers
and musicians. As results from one search bring new names and places
to light, her searches and analytical work will be altered to include
the new names and the new pieces of music. Among the sub-topics arising
in this research are those of racial boundaries in various locales,
exoticism as an element of popular taste, the transmission of stylistic
elements among musical communities, and compositional practice and
intentions of the various individuals identified in this research.
There are several indicators that this phenomenon occurred in the works
of Ellington, Ravel, and Stravinsky, to name but a few, but the larger
study simply cannot be done today. One must be blessed with an unusually
good musical memory and plenty of time for laborious analysis conducted
piece by piece in order to deal with these concepts.
We turn now to the social scientific
scenario, which concerns a team
of scholars in several nations working in the field of international
relations, in particular the questions of the effects of the General
Agreement on Tariffs and Trade and the World
Trade Organization on
bi- and multi-lateral relationships among nations in the period 1950-2010
and the growth of local and national economies in developing nations
in the same period. Among the subordinate questions to be addressed
by this team’s research, are questions of the effects of information
on decisions making. A particular focus will be a select sub-set of
developing nations from each area of the world.
Among the research resources needed for this line of research are many
archival collections of papers from government agencies, non-governmental
organizations, protest groups and labor unions. Government documents,
many of which were issued only on the web, will be needed too, along
with the papers of individuals involved in various of the organizations
and movements. The team will also need a host of statistics in paper
and machine-readable forms gathered by a range of number hunters, from
individual consultants and companies, to national agencies, to the
United Nations. All source materials will be digitized, searchable,
and, for the quantitative data, available for extraction, manipulation,
and analysis.
In order to gather this array of research resources, the team will
need to include librarians, information brokers, government document
curators, and social science data consultants. Source materials will
be drawn from many repositories around the world and installed on a
server for the whole team to use, regardless of their locations. As
data flow in, they will need to be analyzed, then correlated with the
news reports, public statements, and archival records. Some imaginative
work will be necessary too in organizing the documents so they can
be seen in chronological and topical orders as well as arranged by
proposition. The work in this scenario cannot be undertaken on the
grand scale suggested, primarily because the source materials are essentially
unavailable for digitization and thus for analysis. We turn now to
a scenario in the sciences.
The researcher in our scientific
scenario is an ichthyologist seeking
to understand the factors in the depletion of native oysters over the
course of the past 50 years. Of concern are parasites, diseases, and
the biochemistry of disease organisms. At issue as well are pollution,
exploitation, changing water salinity and silting. Geographical and
mathematical models will have to be developed to describe past, and
to predict future, trends. The native oyster in question is Ostrea
edulis, distributed naturally around the British Isles, the North Sea,
the Mediterranean, and the Black Sea; it has been exported in seed
stocks to North America, Australasia, and Japan.
Because of the wide distribution of this oyster over the past centuries,
scientific and trade information about this oyster has been generated
in numerous languages, in peer-reviewed journals, government documents
and in fisheries’ publications. It is known that recent articles
about oyster parasites and diseases are distributed across numerous
journals covering the life sciences, but identifying them will be difficult
due to variance in local naming conventions for the oysters and their
diseases. Local records of estuarine salinity, perturbations of local
estuaries’ banks and floors, and weather conditions will be needed,
as well as reference to the surrounding land and marine topographies.
The roles of improved methods of packing and transportation will be
factored in the problem, as will the inadvertent distribution of exotic
species, competitors, parasites, and diseases in the holds of ships.
Geographical information systems and mathematical modeling will be
employed to analyze the extent of oyster stocks, harvesting, predation,
farming, and disease vectors, leading to a study of oyster epidemiology.
Each of these scenarios is based on research directions and methodological
developments observable now. Each requires access to enormous amounts
of information, only some of it in the published literature, ranging
from the most popular to the most arcane. Each project features roles
for technical and subject librarians to gather, analyze, and model
the information. Not one of these projects can be completed without
recourse to network-based communications and information as well as
to high-performance computing. And all of them will require digitization
of existing data. In the future, such research scenarios will be commonplace,
I predict. The main point here is that we must constantly and continuously
understand what our readers will require, anticipating as best we can
their needs.
The implications for the great research libraries are plain. Applications
of digital methods and global communications have changed and are changing
the nature of research in nearly every discipline. Not a single one
of our functional areas will be left unchanged as we adapt. Most significantly,
our collection development and preservation programs need to accommodate
the new information sources and the new formats. And it is abundantly
clear from these scenarios that the additional means, the tools, agents,
and methods, for broad and deep discovery and retrieval of distributed
information resources will be essential. While on the one hand, the
Internet and the global telecommunications systems have liberated us
from our local precincts, it is true as well that they are making us
more dependent upon other information sources in order to satisfy the
imaginations of our scholars.
For those doubters among you, let me call your attention to the incredible
generative effects the Thesaurus Linguae Graecae had on classics scholarship
in the decade of the 1980s. Think as well about the impact readily
available quantitative data on human behavior has had on the course
of the social sciences and on the practice of politics. And what would
our scientists in particle physics, almost all of the life sciences,
and many fields of engineering have done to pursue their research without
access to computers and to the enormous journal literature in these
fields? Virtually all disciplines now produce results heavily, and
in some cases entirely, dependent upon information technology to make
their findings known and available.
Publishing of all sorts is now heavily dependent upon information technology
and, ironically, there are more published items available and more
or less accessible in printed form than ever before. In addition, for
some crucial scholarly publications, for example the BMJ and
the European Sociological Review, the Internet editions are the most complete versions,
supplanting the primacy of the parallel print editions. And therein
lies one of the great quandaries for the great research libraries:
how to represent all of the records of civilizations’ marches,
whether in our individual collections or collectively, perhaps even
in some coordinated fashion, among our universities. From our perspective,
civilizations do not decline; instead, as they proceed, they produce
more documentary confusion.
It is also plain from these scenarios that libraries in general, and
the great research libraries as leaders in the field, must continue
and, indeed, expand our their efforts to conserve the
artifacts bearing information already in our their care. To do so,
we need to find ways
to re-engineer conservation methods and practices now using craft procedures.
Mass deacidification for acid-laden wood pulp paper, better methods
for dealing with coated paper stock, and widespread adoption of paper-splitting
and reinforcement processes— such as those pioneered by Ernst
Becker at the Zentrum für Bucherhaltung in Leipzig – are
examples of approaches that, if applied at a great many libraries,
would preserve
paper artifacts now more or less doomed to self-destruction.
And presuming that we can develop trustworthy methods
of operating digital archives, a subject that which will arise later,
we should turn to mass digitization, in place of our microfilming projects.
For this to work, we will need some new tools, ones which industrialize
what is now a labor-intensive process. Because we must preserve and
conserve our collections, the great research libraries should take
the lead to develop the new methods to safely accomplish mass digitization.
We need also to purchase services or machines on the marketplace, but
only if they are efficacious, affordable, and do not harm our books.
The wisdom of the monks, scholars, and rulers who first
established universities and then their libraries was not misplaced
with regard to the purposes of these institutions. Any objective, scientific
view of the research product of the great universities would show qualitative
differences when contrasted with other sorts of universities and libraries.
The number and types of connections made possible by large, well organized,
and deeper collections result in qualitatively different research products.
Libraries with great collections make qualitative differences in the
types of research possible in their areas of specialization and across
those areas. This assertion applies to the great research libraries
among the national and public libraries equally well.
There is a corollary to this notion. Great
research libraries are developed
by staff - highly qualified curators, catalogers, arrangers, and others,
whose expertise is applied both to the gathering and organizing of
research resources and to the provision of those resources to scholars.
They contribute significantly to the research possibilities and opportunities
of the great research libraries. Great research libraries afford subject,
genre, language, and technical specialists in sufficient numbers and,
frankly, provide these very special people with working environments
in which their expertise and often peculiar psychologies can be put
to good works.
Furthermore, the three scenarios presented earlier illustrate the growing
dependency of researchers on a mix of digital and traditional information
resources, as well as some new research methodologies requiring computers.
However, and just as in the past, scholars at the forefront of their
disciplines hereafter will likely be caught up simultaneously in benefiting
from – as well as requiring – extensive collections of
information. Great research library collections before the Internet
age stimulated and suggested lines of research, in part by patterns
and leads emerging from the massive collocation of related material
and in part by the work of catalogers, arrangers, and bibliographers
setting meta-information in place and deliberately seeking connections
and occasionally finding contradictions. As a matter of course, scholars
conducting research have made known to bibliographers their requirements
for books, manuscripts and archival material. Canny bookmen have contributed
by keeping their own eyes open for unusual material that might be offered
to libraries to support collection trends already started. This virtuous
tangle of requirements, sources, scholars, and librarians has continued
in the Internet age.
The great research libraries must continue to add to their collections
of unusual material. As such collections grow, so do the number of
planned and systematic opportunities for research. Moreover, adding
unica and uncommon items to already large special collections improves
the chances for serendipitous discoveries. The lively minds working
in our libraries make connections intuitively, as much as from our
cataloging, and follow veins of ideas in mining our stacks for the
subjective glory
holes of gold, different for each individual, but precious for all
nonetheless.
At the Bodley, as reported in its annual reports and press releases,
the continued growth of the collections of medieval manuscripts – while
simultaneously adding modern political and literary papers, music manuscripts,
maps, and Orientalia – mark the progress of this great research
library.
We should note the addition of the late 11th- or early
12-century
Islamic scientific manuscript, “The Book of Strange
Arts and Visual Delights,” which includes a previously unknown
set of maps. This important manuscript is especially auspicious in
this 400th year,
as is the acquisition of the manuscript of the “Hebrides Overture” of
Felix Mendelssohn for the already extensive archive of Mendelssohniana.
Collecting at this intensity of rarities is a mark of great research
libraries. The assembling of the Sackler Library is another sign of
this great library looking to the future.
Selecting a New World sibling library not quite at random, to illustrate
further the point, we have acquired at Stanford in the past decade
the archive of the iconic Beat poet Allen
Ginsberg, the papers of the
Southern Pacific Railway and the Apple
Corporation,
the archive of the polymath engineer and architect R. Buckminster Fuller,
and significant
additions to our archival holdings on John
Steinbeck and
William
Saroyan.
Also residing in the Engelbart collection at Stanford is the 1964 prototype
pointing device, the world’s
first mouse.
And to counter those who might believe that the collecting instincts
of dedicated, even obsessive, individuals will never or rarely direct
them to digital materials, we have examples at Stanford of digital
items coming along with traditional materials, as well as one example
of a collection entirely devoted to computer and video games, the Cabrinety
Collection of software, hardware, and literature.
Remember as well that the Internet Archive is a collection of Internet
sites collected by and for its patron, Brewster Kahle. No doubt there
will be many other such examples coming to us and, as in the past,
one function of the great research libraries will be to provide aid,
comfort, and fellow traveler-ship to individual collectors, including
those most interested in digital objects.
Lest some mistake my theme and belief that great collections are the
foundations of great research libraries for a paean solely to rare
books, manuscripts, and archives, let me hasten to add that great research
libraries also amass huge collections of ordinary books, government
documents, films, music, and journals. Just as the cause and effect
tangle is typical with regard to scholars and libraries and services,
so are there yin and yang complementarities of special and ordinary
collections in the great research libraries. Adding digital objects
to the mix augments the networks of ideas that span the common as well
as the obscure, awaiting invigoration by inquiring minds, exploring
our physical and virtual collections. Unusual, special, rare collections
need auras of general collections, quickly and easily available, to
be fully exploited. And conversely, great, large, general collections,
accumulated over time, become more and more “special” as
they become deeper, wider, more multi-lingual, and incorporating more
points of view, records of more controversies, and successive layers
of cultural engagement.
It is not my intention, however, to deny or belittle the development
of research collections in other sorts of libraries. In fact, I believe
that each library, no matter how small, has contributions to make to
the development of the logical, global research library, that greatest
construction of them all. Regardless of the motivation to develop a
collection which might serve research purposes, whether one’s
collection policies are devoted to local history, to further elaborating
a collection assembled by a donor, or even to focus on a single specialty,
we can contemplate providing intellectual access to all such collections
on a global scale.
I do not advocate that all libraries attempt to become research libraries.
Indeed, I do not believe that most research libraries should pretend
that they can become great research libraries, because there are too
many examples of the diffusion and resulting waste of local and national
resources to lowest common denominator acquisitions programs, altogether
too plain and too vanilla, contributing too little of unique or special
substance. Rather, emphasis should be upon the ways that any and all
libraries can contribute uniquely to the commonwealth of knowledge.
So, now some thoughts on other roles of libraries in the future…
What has the Internet Age meant
to libraries and particularly to the great research libraries? What
will the great research libraries undertake
because of it? The transformation and the possibilities can
be seen in several functional perspectives: access, distribution, analysis,
collecting new genres, preserving digital information.
The possibilities for libraries to provide intellectual access
to their collections have widened considerably. From the
1970s, library staff
have been able to search on-line union catalogs, such as those of OCLC
and RLG, the two notable survivors of the shared cataloging stage of
the use of information technology by libraries. Now we present
our catalogs of our holdings via browser interfaces, sometimes hyperlinking
from catalog records to digital editions as in this example of a link
from an OPAC on the left to the European Journal of Biochemistry. In
addition, we provide access via on-line abstracting and indexing services
to the unanalyzed contents of journals and anthologies in our collections.
Some of us have begun to provide synthetic guides to the
literatures of various disciplines, usually prompted by twinned
forces: the desire
to make readers more self-sufficient in their information discovery
and retrieval activities and the need to provide guidance to readers,
who are novices in disciplines not their own. We are now seeing an
increase in the amount of hyperlinking from cited references to the
sources cited in scholarly communication and similarly from abstracting
and indexing services to the Internet editions of the sources they
cover. Developing this web of hyperlinking will contribute to the next
great phase of the Internet, that of the Semantic Web, by increasing
the amount and types of meta-data. We in the great research libraries
can contribute to the development and operation of the Semantic Web
and should cheerfully contribute metadata for our holdings. We should
develop taxonomies and ontologies to permit cross-format and multiple-genre
relationships to be extracted. Also, we should report to readers, on
demand, what others interested in their topics or research resources
are reading.
There are other developments in which publishers, librarians,
and information technologist working together have created
narrowly specialized digital
libraries of information and information services for widely dispersed
communities of scholars. Such “knowledge environments” provide
access to journal articles, newly commissioned perspectives and reviews,
protocols and guidelines for conducting research, alerting and customization
services, and often unique, methods for displaying and navigating pertinent
information services. In addition, these knowledge environments offer
communication services in the form of on-line open and moderated forums
in addition to personal and institutional directories. Once there are
a great many knowledge environments, we will need to provide the means
to perform discovery and retrieval functions among them as well as
support cross-disciplinary twigging and cloning so characteristic of
truly healthy evolution of scholarly enterprises. On the screens now
are examples from the BoneKEy knowledge environment focused on bone
and mineral research in medicine and another on cellular signal transduction
with its unique connections map showing flow charts of intra-cellular
biochemical processes with hyperlinks to explanatory information from
each information “blob.”
Despite the wealth of access possible through various network
based metadata sources, performing a literature search across
all – or
even some – of them is a daunting task. One of the major tasks
facing research libraries is the development of discovery and retrieval
engines that will take a single search argument and apply it across
a range of metadata sources, ideally a range set by the searcher. There
are some early versions of such a research tool. FlashPoint from
the library at Los Alamos National Laboratory searches BIOSIS, the
various
sections of the ISI citation indices, Engineering Index, Inspec and
the e-Print arXiv. Another, the search
engine at HighWire Press, simultaneously
searches all of Medline and about one million articles in over 340
journals. Here also is the HighWire
Topic Map. Other search engines
are in development, but we will go through many generations of increasingly
sophisticated versions before we satisfy our readers.
In addition, there needs to be much creative work on the
effective display of the “hits” retrieved. Eventually, we will want
that sort of discovery and retrieval engine to seek words and phrases
in complete texts as well as meta-data.
The great
New Yorker cartoon defines in the broadest terms
the effects of the Internet on publishing. Anyone with a
computer and a link to
the Internet can become a publisher. The mass of popular “publishing” presents
a problem to archivists and those scholars of contemporary culture
who want to preserve at least a sample of popular discourse from our
age. In the arena of serious publishing, the Internet has created a
maelstrom of change, threatening some traditional players, while simultaneously
empowering others. Certainly scholarly and serious journals have been
transformed, with many new features like hyperlinking to cited references, “prospective
citations,” multiple resolutions of images, and opportunities
to communicate directly with authors. And we are seeing the beginning
of the transformation of the scholarly monograph along the same lines. “Classic” texts,
from the ancient philosophers, to the Latin and Greek Church fathers,
to Marshall McCluhan can now be read on the Internet, sometimes free
of cost, sometimes not. Printed music is coming to the web. These digitized
editions in their simplest form, such as the 4,000 titles in Project
Gutenberg, make out-of-copyright titles widely available, but encoded
digital texts allow new forms of research, involving comparisons at
various levels of literality and meaning. Examples include the Miguel
Cervantes Virtual Library, the Electronic Text Center at the University
of Virginia, The British National Corpus, and the Forced
Migration Online, both at Oxford.
Libraries have new possibilities for distribution of their
holdings too. Lately, as scanners with better resolution
and more efficient
handling of books have become available, some notable experiments in
digital interlibrary loan of whole books have been undertaken. We are
beginning to see more automated scanning devices that deal gently with
books as objects while providing accurate conversion of the captured
images. These new devices promise to make scanning of books and documents
less costly and faster, thus opening the possibilities not just for
lending books on-line, but also for the analysis of the texts.
E-book distributors like Ebrary offer on-line access for library patrons
and readers everywhere to currently published books, albeit in small
numbers as yet. They too offer good searching of the texts for words
and phrases and other useful functions. Even in this early stage, searching
a collection of books for words and phrases has begun to provide the
same magnitude of increase of benefits to our readers, as did the retrospective
conversion of our card catalogs, despite the earlier caviling of Nicholson
Baker.
One should not neglect to mention the factor of convenience
of access to on-line information resources resulting
from the Internet Age. For
our local readers and staff, the provision of on-line access 7 x 24
has meant that procrastination can rise to new heights of risk, delaying
until the wee hours that bit of research and writing due soon after
the dawn. But, it has also meant that we can operate 24 hour libraries
without burdening staff to remain in the physical facilities in the
same wee hours.
Digitizing some of the treasures of the great research libraries
is well underway, but more should be done. Various projects
of the Bodleian
itself are contributing digital
facsimiles of early manuscripts from
the collections of Oxford’s colleges and from the Bodley itself.
Other great
libraries around the world have begun projects such as
these to demonstrate the possibilities for research and teaching. Where
formerly we might make a microfilm master on demand, with roughly the
same investment we can produce a digital copy and make
it globally available. Some of us have concerns about vitiating
our institutions’ cultural
patrimony, but, those concerns notwithstanding, it is clear that the
possibilities for assembling virtual collections of materials - related
by provenance, by subject, by author, and by many other characteristics
- could contribute to much more efficient scholarship than our reliance
on physical copies allowed in the past.
Contemplate digitizing ALL of the items in the collections
of one of the great research libraries and then making as
much of the digitized
material as possible available to the world. Yes, of course, there
are copyright issues to be solved, but imagine the effects of making
not thousands, but millions of books and perhaps tens or hundreds of
millions of journal and newspaper articles available for searching
and retrieval on-line. Is it too much to say that the results would
be a new Enlightenment? Maybe we will see such a project underway one
of these days.
I have mentioned new forms of analysis in the research
scenarios and now want to point out the additional requirements for
libraries to make these new methods available to our researchers. The
various sorts of search and comparison functions necessary for advanced
textual research methods require libraries to mount the texts and provide
data searching and manipulation software. Encoded texts, whether using
SGML or XML and their variants, provide more opportunities for advanced
research. Many publishers, e-book distributors, and libraries are providing
texts and search engines. However, occasionally, texts must be entered
into new data formats to enable special sorts of examinations. In these
cases, too, the great research libraries will be called upon at least
to supply the desired texts to the research teams, and often to adopt
the new data formats. The full repertory of such transformation of
texts, prototyping, and adoption of new methods is probably without
practical limit. It is important for librarians to serve on the research
teams, contributing the development of the new methods.
New analytical methods are constantly emerging. One of the least developed
is that of “sonification” of
data for navigation of large and complex data sets by sound. A research
project at the Stanford
Humanities Lab has treated oceanographic data and stock market performance
in gross and by individual company equities. Will this research lead
anywhere? I am not sure. But I am sure we need to be cognizant of such
exploration. Another area, now in heavy demand, is analysis of social
scientific data, much of it generated by government agencies. All the
great research libraries offer statistical
data in digital form, but
using it requires specialized assistance. As the use of social scientific
data increases, the great research libraries will develop web-based
interfaces so that readers and researchers can get to the data they
need without assistance by library specialists, who will then be available
for more specialized consulting services. To conclude this section
on analytical tools, I will digress significantly to a
demonstration
of the advanced use of geographical information systems to better understand
history – in effect to see it in new ways – and ultimately
to better represent our times.
The work
of David Rumsey, a private collector of maps, atlases, and
related material is illustrative of some new analytical methods, in
his case ones utilizing a variety of new Geographical Information Systems
(GIS) and some incremental advances he has stimulated in information
management.
Rumsey has been collecting for a long time, but over the past few years
he saw and then adapted the possibilities offered by information technology
to make his collection accessible, to add information to digital versions
of his maps, and to produce a variety of new views of those digital
versions.
Rumsey uses
Luna Imaging’s Insight Software and various GIS software.
One can use three different means to interact with the digital
images and data he provides. An ordinary web browser using
Java scripts can be employed. An Insight Java client can
be downloaded with advanced functions. Or, one can download
a GIS browser with an applet.
David has associated himself with Berkeley, Stanford, the American
Antiquarian Society, and three commercial interests, Luna Imaging,
Telemorphic Software, and AMICO, the art image distribution agency
for a number of important museums. Here is a screen showing the Arrowsmith
map of 1844 and the numerous functions of Luna’s InSight software.
Rumsey shares his collection in many ways: He has contributed 6500
of his metadata records, with hot links to his images, to UC Berkeley’s
online catalog.
Rumsey makes his maps available through many tools and intermediaries:
Google, the Open Archive Initiative, Mellon’s OAIster, E.S.R.I.’s
shared Geography Network, and the Electronic Cultural Atlas Initiative.
Let us look at some GIS functions that Rumsey is making possible for
researchers. Here is Colton’s 1836 Map of New York City, georectified.
Then it is overlaid with modern street and shoreline data for the New
York City metropolitan area.
Zoom in to the area around Central
Park
We can see changes to the shoreline on the left, landfill for the Henry
Hudson Parkway, and see changes in Central Park – notably, the
placement of the park itself in 1856 and the moving of the reservoir
north, etc.
We can compare historical maps to each other more effectively in GIS.
Here, the Colton 1836 NYC again.
And here a map physically twice the size of the Colton, the Dripps
Map of New York, 1852.
When both are shown in GIS, the Dripps map is now about 1/6 the size
of the Colton, because it is on a much larger scale.
Zoom in to lower Manhattan with the 1836 Colton
Then overlay the 1852 Dripps map as a transparency.
Zoom to Tompkins Square on the 1836 Colton
Then overlay the 1852 Dripps map of the same area and see
changes over 16 years of development.
In1802, Thomas Jefferson sent Lewis & Clark to
explore the far reaches of the Lousiana Purchase. The Bodleian Library
was only 200 years old then.
Rumsey is using GIS to allow
enhanced interpretation of the historical maps as well as
the ability to combine
them with modern geospatial data. This method permits us to more accurately
trace
Lewis and Clark’s route of discovery.
Here is the 1814 published map of
Lewis and Clark’s Journey to
the Pacific, in its original form, before georectification
Here is the same map after georectification.
The amount of change is not too bad considering that Lewis and Clark
were using inaccurate
clocks to determine longitude and lots of estimating by eye up and
down mountain ranges and valleys. This example took about 40 known
points to georectify.
Now we can combine the old map with modern map data
showing major roads, some cities, and state boundaries.
And we can even add vector data that
shows all the Lewis and Clark camp sites (in yellow).
GIS can also be used to create interesting mosaic maps – here,
the Lewis and Clark
Map of 1814 is shown with a 30 mile buffer on each
side of the trail route, which then blends into an 1879 US General
Land Office Map of the same area, which then blends into the US 1970
National Atlas Map, and finally the mosaic is bordered on all four
sides by current satellite imagery of the area. The following slides
show how the blending works
First, showing the blend with the 1814
Lewis and Clark Map gradually
becoming more prominent.
Next, gradually showing how the 1879 US General Land Office
maps are
involved.
Finally, swiping in the current Satellite
imagery. [Click for map
blending and overlay animation]
On the stand over there is a photo of the blended map for you to inspect
more closely after this presentation is concluded.
Rumsey is also working with 3D GIS. Here is a 1926
maritime Chart of San Francisco Bay showing depths with numbers on the chart
surface This method allows us to perceive the shape of the surface
of the Earth obscured by water. .
Here is a modern bathymetric model of the same area of the bay with
depths shown in GIS in exaggerated fashion.
We can combine the two and see the depths shown on the 1926 chart using
3D GIS.
Now, zooming in the Golden Gate straights we can see the trench gouged
out by tidal action.
We change our viewpoint as if we were hovering over Oakland, looking
west toward the Golden Gate.
Zoom in to Yerba Buena Island, shown on the original map in 1926 as
a small hilly one, then in the combined data view here we can see the
land fill extending the island north for the 1939 World’s Fair,
now known as Treasure Island
Another example of 3D GIS – taking this 1915
Wall Map of San Francisco by Chevalier and
combining it with current Digital Elevation Models (DEM’s) from
the United States Geological Service.
We “Shrink wrap” the 1915 SF Map using 3D GIS.
zooming in and turning, so that the peaks to the south are in the top
of the screen.
We can change the source of illumination – [move through the
next slides quickly as in animation to show the changes in illumination.]
Note the graphic in the lower right corner showing azimuth and altitude
of the sun’s rays We can change the illumination … not
add fog …
Now let’s Overlay modern streets on the 1915 map
Here are the 1915 streets so you can compare growth of the city infrastructure.
These functions allow us to see the landscape in varying conditions
of sunlight, and to plot the growth and change in the layout of streets.
Such representations of reality help some better understand their surroundings.
David Rumsey is remarkable, because he is investing heavily in making
his maps available over the Internet in a variety of ways. His motive
is to excite others about the beauty and historical interest of his
map collection, but also to allow them to be used in digital editions
he has created.
Now we turn to new genres and literatures.
Hypertext and hyperlinking provide report writers and fiction writers
with new possibilities. Certainly we have seen the use by undergraduates
of our facilities in inserting images, charts, graphs, tables, moving
pictures, and sounds in reports and projects. Because an increasing
proportion of these reports involve multiple media and texts, they
often cannot be delivered on paper, but must be delivered to their
teachers in digital form. And as some of these students become professional
scholars, we will see more such mixed media reports and articles in
scholarly communication. A few examples of this phenomenon were seen
as early as June 1995 in Science Magazine,
and other journals, like the Molecular Biology of the Cell, followed
suit shortly thereafter.
In the example on the left … the example on the right … entitled … Babbling.
However, in the realm of creative writing, hypertext has
provided some new possibilities for authors, not only to
tell stories, but also to
engage readers in making choices in the directions a story might take.
Here are a few examples of how hypertext fiction can work from Richard
Holeton’s “Figurski
at Findhorn on Acid.” The new
forms of expression are almost certainly extrapolating from video and
computer games as well. Multiple-user Dungeons and Dragons provide
directed environments for playing out tales of battles, conquest, and
general mayhem. Similarly, other gaming environments involve sports,
historic settings, exploration of exotic places, and just plain gambling.
All of these have the potential to inform and to increase the expressive
possibilities for writers of all genres.
Some academics read more into this. Witness this excerpt from a prospectus by
a team of scholars working on the history of video and computer games. “ A spectrum of theorists …proclaim that we are
on the verge of a new – albeit post-human – renaissance,
an era in which the human being becomes seamlessly articulated with
the intelligent machine, a condition in which there are no demarcations
between bodily existence and computer simulation, between cybernetic
mechanism and biological organism.” Think about the consequent
opportunities for new items for our collections!
That there may not yet be a “classic” in the
hypertext genre is irrelevant to the responsibility the great
research libraries
must bear. Here, at the birth of so many new genres of expression and
communication, we must collect examples of each, perhaps widely, and
then we must preserve them so that scholars, students, and pundits
in the next generations might examine and evaluate them for the history
of our present. We cannot distinguish the quick from the stillborn
in and among these genres.
The prospect of collecting the new genres, ones like hypertext fiction
and video games, raises one more function thrust upon the great research
libraries from their early days, that of preserving ideas and expressions,
data and doggerel born digital as well as digitized. To rehearse local
history, remember that Thomas Bodley stepped into the breach caused
by the gradual disappearance of the manuscripts given by Humfrey, Duke
of Gloucester, to Oxford between 1435 and 1444. Bodley could not replace
the manuscripts of either Duke Humfrey or the earlier benefactor, Bishop
Cobham, and Oxford has recovered only a few of them over all these
centuries. Thomas Bodley could and did, however, set in place a library
tradition, with suitably designed and constructed facilities, involving
professional, full-time librarians. He established rules that have
preserved the collections for these many generations of readers and
stimulated the development of the collections by gifts of his own books
and those of others. By negotiating the 1610 agreement with the Stationers
Company for what amounted to copyright deposit, Bodley set the scene
for the development of many great research libraries.
It is now time to fill another breach, that of the disappearance
of ideas and expressions born, published, and distributed digitally. There
are too many examples of important information gathered and published
in the past few decades simply disappearing, either because no effort
was made to collect and preserve it or because the technology storing
it or necessary to read it has not been maintained. We thus lack images
of the Earth from the NASA Landsat photograph collection from the earliest
days of the orbiting cameras. Census tapes from the 1960s are unreadable.
Satellite observations of Brazil in the 1970s, critical for establishing
a time-line of changes in the Amazon basin, are also lost on the now
obsolete tapes to which they were written. It is not only the data
that are fragile, but also the rapid passing of generations of software
and hardware that will leave us unable to provide access to those digital
objects. Clearly the leadership of the British Library and the establishment
of the Digital Preservation Coalition here are significant steps in
the creation of trusted digital archives. The National
Digital Information Infrastructure and Preservation Program of the Library of Congress,
now concluding a year of intensive planning, should lead to support
for some prototypes of digital archives in the U.S. Work at the National
Library of Australia in the PANDORA project is notable too.
Stewart
Brand,
founder, editor & publisher of the Whole Earth Catalog
and co-chair of the Long Now Foundation, has said that “exercise
is the great preserver [of data].”
It is that notion which underlies the operations of the network caching
software, known as LOCKSS, which many of you have beta-tested. LOCKSS is an acronym for Lots Of Copies Keep Stuff Safe. Brand and many in
this room today have good reason to fret that unless we in the great
research libraries begin to preserve digital objects, we may be creating
what our successors will label a digital dark age.
Another breakthrough is the recent announcement that the Koninklijke
Bibliotheek, the National Library of the Netherlands, will become
the first official digital archive for all 1,500 Elsevier Science journals
amounting to over 7 terabytes of data. The Elsevier decision to allow
the Koninklijke Bibliotheek to function as its digital archive is recognition
that libraries will function in their traditional role even when the
media of the objects under care change. More will follow suit, I am
sure, especially when some of the prototype
digital archives now contemplated
have proven satisfactory.
[The
Dark Side]
Optimistic prognostication about what great research libraries ought
to do and logical extensions from present programs and operations notwithstanding,
we must face the dark side of the Internet Age.
It threatens to inhibit our future and our abilities to deliver information,
ideas, and expression
to our readers now and for a very long time to come The continued imbalance
of values and valuation between the originators of ideas and expression
and certain irresponsible publishers is troubling. Scientists and scholars
consider the reports of their work to be important for many reasons,
including indirect monetary gain. Certain publishers, on the other
hand, see such reports as information to be treated as a commodity.
There is a resulting clash of value systems involved here, but not
many librarians have recognized that in holding the purse strings,
they hold what may be one of the keys to improving the present imbalance
to one more favorable to scholars and their institutions. Librarians
in the great research libraries must retain their selectivity and avoid
re-allocating funds– from the humanities and social sciences – to
pay for the excessively expensive journals in the sciences, technology,
and medicine. And in particular, the great research libraries need
to recognize the responsible publishers among the herd and suggest
that scholars send their manuscripts to them.
Librarians who have allowed themselves to be seduced by the “big
deals” offered by many of the for-profit scholarly publishers
find themselves stuck in long-term contracts and increasing expenditures
for a single publisher’s products which spending then results
in decreased spending for other books and serials. Having succumbed
to the blandishments of a lot more information for seemingly small
increases in cost, some librarians are now discovering that much of
the information added to their original careful selections is little
used by their readers. Cancellations are forbidden or severely restricted
by the long-term deals, so responsiveness to one’s primary clientele – faculty
and students at one’s home institution – is sacrificed
as well.
The most grievous difficulties are arising, in my view, in new laws,
amendment to old laws, changed regulations, and cunningly developed
international treaties, all limiting the rights of citizens in democracies
to freely read, otherwise observe, and finally make use for private
purposes copyrighted intellectual property. In the U.S. there has been
a pitched battle, fought mainly by the Disney
Company and
other popular media barons as well as the Association of American Publishers.
The
results have been the provisions of the Digital Millennium Copyright
Act and various attacks on librarians. The World Intellectual Property
Organization and the various attempts to “harmonize” approaches
to European
Community and national intellectual property codes are
producing inhibitions to creative work as much as they are rewarding
the actual creators of new ideas and expressions. There seem to be
two central issues in this thicket of protectionism, on one side advocacy
for the rights of commercial interests and, on the other, concern for
the rights of citizens to exchange and use information. The first issue
involves the regulation in the digital environment of who can read
what, where, and for what cost.
Essentially those interested in the presumed rights of companies already
in the publishing game for profit (and lately by controlling access),
as well as keeping high the barriers to entry into the publishing game,
have heavy influence on legislators and legislation. The second issue
is the right of the creators to establish their own means of publishing;
it arises as the creators and their home institutions (universities
and scholarly societies), attempt to regain their own rights to publish.
Such social movements as the Public Library of Science illustrate,
in the irate fervor of its adherents, the furor these issues are creating.
In the U.S., fair use, a doctrine defined in the U.S. code, is under
constant attack by those who believe that every time a digital version
of anything is read, some money needs to be sent to the copyright holder.
And fair uses of media material in academic settings for teaching and
research are objectionable too. There are now bills in
the U.S. Congress making criminal some violations of copyright.
For those who have even passing interest and engagement in this topic,
the work of professor Larry
Lessig
is essential. Lessig has written, in his book “The Future of
Ideas”, that the way we as a
society resolve issues of access to digital content, whether networked
or not, "will determine what the 'free' means in our self-congratulatory
claim that we are now, and will always be, a 'free society.'"
Lessig is notable also for establishing the “Creative Commons,”
a site providing authors with a set of tools and advice on making their
creative works available with terms they specify rather than those
of commercial intermediaries.
The Chilling Effects Clearinghouse is
one among several initiatives to address the problem of the Dark Side’s
intellectual property issues from the viewpoints of citizen-readers.
Chilling Effects focuses
on the U.S., but there are similar concerns elsewhere. For example,
there is a site by Paul Burton here in the U.K. devoted to the question
of control of the Internet with examples and issues drawn from all
over the world. Running out of the University of Leeds is another site
devoted to cyber-rights and cyber-liberties.
Librarians and law school professors are not alone in this fight. In
the U.S. there is a new digital
consumer organization seeking to develop
a grass roots coalition to lobby against the well-funded, well-entrenched
representatives of the Disney company and their industry . And one
Congressman, Rick Boucher of Virginia, has taken up fair use as one
of a series
of positions protecting the interest of consumers, of ordinary citizens.
It seems that the work of the Joint Information Systems
Committee and
the Publishers Association in developing the Guidelines for Fair Dealing
in an Electronic Environment characterizes a more collegial and straightforward
attempt to deal with similar issues. Those guidelines are not without
tension and stress, I know, and the Publishers Association website
features its rather protectionist 1996 statement on The Use of Digitised
Copyright Works in Libraries but not the jointly agreed guidelines.
On the other hand, the UK Commission on Intellectual Property Rights
with its focus on intellectual property regimes and their impacts on
poor and developing nations is an example of an enlightened government
effort to improve access to information in poorer nations while improving
the observation of intellectual property rights. A segment of the statement
of Libraries and Archives Copyright Alliance on
copyright in the digital age strikes a balance in the copyright issues
for many of us. I quote: “However,
LACA maintains that overprotection of copyright could threaten democratic
traditions and impact on social justice principles by unreasonably
restricting access to information and knowledge. If copyright protection
is too strong, competition and innovation is restricted and creativity
is stifled.” Close quote.
The role of the great research libraries in this arena is to advocate
for the rights of our readers now and in the future. Laws and regulations
impinging upon our functions as custodians of culture need to be identified
as such and we in the great and smaller research libraries need to
join the coalitions of those on our side of the issues, the side of
fair use and fair dealing for private, educational, and research purposes
of information.
The World Wildlife Foundation issued
a report in July 2002 projecting that the Earth’s biological
assets could be exhausted by the year 2050. Overpopulation, overfishing,
pollution, and many other factors
suggest that we need another planet to
occupy, because we are rapidly consuming this one. If this is the case,
the role of the great research
libraries in capturing as much of the key documentation of human activity,
especially in the industrialization of agriculture as well as the clinical,
nutritional, and pharmaceutical “advances” effectively
increasing the number of humans alive at once, could provide the few
survivors of the looming disaster with valuable information crucial
to the survival of our species and culture.
[CONCLUSION]
In the few years preceding the formal opening of this library in November
of 1602, Thomas Bodley could hardly have conceived of the extent of
change libraries were to enjoy and endure over the centuries until
this day. In Bodley’s lifetime, books were only slightly less
rare and less talismanic of intellectual pursuits than they had been
at the time of Duke Humfrey’s gifts of books to Oxford 150 years
earlier. The chains that bound the books in
Humfrey’s and Bodley’s
libraries in their earliest incarnations were tangible reminders of
the status of books as treasures, to be consulted in the earliest days
only by senior members of the faculty. Oxford’s earliest records
about its books mention several trends which continue to this day:
solicitation for books and funds for the university’s library
collections and facilities; inconsistent attention to support for new
acquisitions and for the upkeep of the library; but, on the other hand,
sporadic periods of deep concern and conscientious engagement by university
officers to address and ameliorate the states of affairs of the Bodleian
Library.
The Bodleian has been joined in this wonderful status by
a few other great research libraries and a slightly more
numerous lot of research
libraries with a few great collections each. The fact that these other
libraries have arisen in some sense gives comfort and aid to the Bodleian’s
keepers and masters, but could as well lead to the diminution or weakening
of support for the acquisitions and operations, which in the case of
the greatest research libraries are qualitatively different than in
all others.
It is a mistake to jumble all sorts of libraries of all sizes
and conditions as an easy way of coping with changes in economic
fortune, of technology,
and of the nature of research. The convenience of such mélanges
of libraries to policy makers and administrators, people earnestly
trying to deal with recalcitrant problems of asset management, is exacerbated
by the sort of thinking represented by themes like “access instead
of ownership” and by the famous New Yorker cartoon “On
the Internet, no one knows you are a dog.” To the contrary, however,
one can discern dogs on the Internet, great collections do determine
and provide access, and the functions and possibilities of great research
libraries are different from those provided by all other sorts of libraries.
There are only a few themes in this presentation. Most important is
that Deep and broad collecting is essential among the great research
libraries, especially of unusual materials and of new sources and genres
of ideas and expression in digital forms. Research libraries still
need to find, nurture, and take advantage of subject, language, and
technical specialists, some of which will be involved in specifying
and perhaps even inventing new technologies.
Thanks to the advances in information technology and the Internet,
the great research libraries have new opportunities to serve scholars
by new methods of analysis of their holdings, some of which will need
to be digitized, and by distributing access to their holdings globally.
In taking advantage of information technology in these ways, the great
research libraries will be responding to and stimulating developments
in research methodologies as well as the formation and reinforcement
of communities of scholars right around the Earth.
The great research libraries need to collect digital information resources
for the present and the future, and they need to discover ways to protect,
conserve, and make those resources accessible in the face of rapidly
changing technical environments, legal and financial hurdles, with
as yet untested methods. They need to cooperate with one another in
operating digital archives so that there are always multiple copies
of any digital resource, on the grounds that in redundancy lies long
term safety. They need to make accessible digital information resources
to their readers, just as they did printed and manuscript sources of
earlier centuries, in part because exercise of the data ensures its
longevity and in part because use of information for private, educational,
and research purposes is a key factor in the persistence of democracy
itself.
Today, a week after the first anniversary of the terrorist attacks
in America, the Bodleian and all libraries, whether their future is
bright or dim, represent and reflect the best, the highest aspirations
of man. In stark contrast to terrorism, libraries and their staff protect
the freedom to speak, to read, and to think. Long may it be so!
Thank you very much for your patient attention.
Michael A. Keller, University Librarian, Stanford University. A talk for "A Celebration of Libraries" on the 400th Anniversary of the Bodleian Library, Oxford, 19th of September 2002
