Building Digital Libraries: Scanning
Is Still Hip High-Tech
by TERENCE K. HUWE
As a profession, we have navigated some pretty rugged
terrain, particularly if you like to trace the online
"dawn of time" to the late 1970s like I do. When I enrolled
in library school in 1979, 300 baud-rate connections were
cutting-edge. Since then, and in stark contrast to many
friends of mine who work in professions like law or accounting,
I find that I learn how to use at least two new technologies
per year. Of course, this is standard procedure in our
profession, and our long-term experience provides plenty
of fodder for that perennial favorite at the Computers
in Libraries conference, Looking at Dead & Emerging
So it's all the more fun to take a closer look at what's
new in the ultra-hip, ultra-cutting-edge world of scanning.
Scanning is far from dead. It remains a fundamental
skill for digital librarians. Our collections, even
when you subtract the percentage of copyrighted material,
still have vast amounts of print treasures that are
waiting to go digital. The richest of us can contract
whole collections to jobbers who will, for a tidy fee,
create searchable PDF documents that fit very nicely
into digital libraries as objects within a persistent
link. Images? Hey, the only real difference between
an image file and a text file is the need to guarantee
a "high rez" archival copy. Both files need metadata,
which can be added to the digital object in fancy repositories,
or simply typed as "meta" links in garden-variety HTML
pages in a no-frills Web environment. Well-funded projects
can be turned around in a matter of weeks, delivered
on a set of CDs, and uploaded forthwith. That's the
fancy way to go.
But not everyone has the wherewithal to farm out scanning
work. Let's face ityou don't always get every
grant you apply for, but that's not a reason to delay
creating digital resources by scanning. We have grown
accustomed to hearing about our largest and most respected
research libraries' latest digitization projects. Some,
like the Library of Congress' American Memory project,
are truly monumental in scope (http://memory.loc.gov).
But the glamour of the monumental can obscure the other
end of the spectrum, where you, as a tech-savvy collection
specialist, can create a modest-sized, manageable digitization
project. This kind of cottage industry can do more to
enhance your reputation as an innovator than many other
activities. And the new generation of scanners makes
it easier than ever to get started.
Scanning's Strategic Value
Nowadays, more and more books, journals, and other
library materials are "born digital," and our jobs are
increasingly becoming devoted to crafting access strategies.
But what about the still-relevant corpus of knowledge
that was not born digital? The Northeast Document Conservation
Center continues to offer a School for Scanning conference
to help professionals learn best practices for conversion
projects, from start to finish. This hands-on approach
is immensely empowering, because it can help you master
the details of OCR, metadata creation, and archival-quality
digital objects. Indeed, the skills associated with
print conversion will remain in demand for some time
to come and are the foundation of digital library development
But a new generation of scanners brings more power
to your hands without so much formal training. These
products are reasonably priced and are much more reliable
than their ancestors, and they enable you to get started
with digitization without depending on grantsmanship.
Here are three mission-critical reasons for re-evaluating
your digitization plans, in light of the new technology.
1. Portability. As is the case with so many evolving
hardware products, scanners have gotten smaller, quicker,
and lighter. This can be handy if you are conducting
research in the field or are just building a database
of articles for personal use. Sometimes it's preferable
to set up a scanning session in an out-of-the-way place
so that you can examine collections of documents, photos,
or ephemera in their "native" habitat.
2. Scalability. Anyone who lived through retrospective
conversions of card catalogs knows what industrial-strength
automation is like: huge, incremental, and slow. However,
a lot of really interesting digital libraries lie waiting
at the other end of the project spectrum: not too big,
but high in value to researchers. For example, my library
has the most complete collection covering the "California
School" of industrial relationsthe cooperative
history of labor and management that grew during World
War II, particularly at the Kaiser shipyards that built
"Liberty ships." Scanning this collection and creating
a digital library could be achieved in months, not years.
3. Marketing. In an era of instantaneous "googling"
and ubiquitous Web information, we need to distinguish
our collections and services by emphasizing their unique
qualities. Many of these unique resources were not born
digital. Digitizing them creates an instantaneous marketing
opportunity to distinguish ourselves, and our library's
imprimatur, from the "semi-decent, sorta OK" world of
the open Web.
Meet the New Scanners
Hewlett Packard's 4670 is a lightweight, flat panel
scanner that you can carry with you into the stacks
or even to remote locations. It's simple to use, and,
in contrast to the bad old days, the OCR software is
superlative, producing clear text copies in one scan.
It will drop a finished scan directly into Microsoft
Word, from which you can create a portable document.
Add a metadata header, and you've got a digital object
in your hands. Locally, we've used this scanner to create
digital images of posters, newspapers, and ephemera
with unique historical value.
Hewlett Packard's unit is a "best in class" entry
that most librarians will like, but there are smaller
units that aren't even designed with the desktop in
mind. The DocuPen, which was evaluated in The New
York Times Circuits section (July 8, 2004, p. E8),
is so compact that you can carry it with you wherever
you go. Moreover, it's designed to help organize and
keep track of scans, along the lines of those less-facile
digital rolodexes. Its weakest point is the resolution
quality, which comes in at 100 to 200 dpi (dots per
inch). This pretty much bars creation of archival-quality
scans, for which 300 dpi is considered the lowest acceptable
number (with 600 dpi preferred).
For the large-scale scanners among us, there's Fujitsu's
ScanSnap FI-5110EOX, which is a moderately priced but
high-powered machine aimed at the records management
market. This scanner will copy 50 pages at a time and
convert them into searchable Adobe PDFs, ready
for uploading. Fujitsu envisions this kind of "almost
industrial strength" scanning as a solution for paper-intensive
industries and services, like law and medicine, where
meticulous records and multiple copies are required.
What About Scanning for Smaller Libraries?
In each of these cases, the scanners aren't being
actively marketed to librarians. There are probably
several reasons, not least of which is the fact that
large funding agencies are still paying for library
digitization projects. The libraries that can handle
large-scale digitization also tend to have sizable staff
complements and can integrate new work flows more easily
than smaller libraries. But those small librariesperhaps
yoursmay have special collections that could be
readily scanned with the new generation of technology.
Ironically, smaller libraries are often hit the hardest
by budget cuts, as anyone in California can attest.
But tight budgets shouldn't create roadblocks for creative
work. Indeed, in times when the government faces shrinking
budgets and firms contemplate outsourcing, creative
thinking is the gateway to survival. Here are three
strategies I can think of to leverage your current collection,
with the new scanners that are cheap and easier to use.
1. Survey the Collection. What stashes of photographs,
memorabilia, or other historical materials do you have?
What printed records do you have that will have historic
value? Recent historical research about the antebellum
South relies heavily on court transcripts to give the
most accurate description of slave life before the Civil
War. Is there something of similar value lying dormant
in your stacks? As information professionals, it is
our job to bring these materials to lightand to
keep them accessible beyond today's fads or scholarly
2. Use Newly Digitized Collections to Be Distinctive.
Library Web sites and blogs, at their best, offer visitors
distinct experiences and unique resources. However,
all too often we stop at the mere description of our
print collections, which doesn't tell the visitor all
that much. Select a small, manageable portion of your
unique materials, scan them, and upload them: This will
make your collection more eye-catching.
3. Start Small, Grow with Time. Success
begets success, nowhere more so than in grant writing.
Oftentimes a tangible resource, like digital collections
in nascent form, can help win really big bucks, especially
for persistent grant writers who establish a track record.
Creating a prototype digital library from scratch may
yield a new revenue stream later on, or it may help
you build a relationship with a consortium, based on
new awareness of what you have to offer.
The Future of Scanning
As always, the human side of digital library development
involves some fairly nitty-gritty worklike scanning.
But cumulative effort sustained in small bits and pieces
over the long term can yield big results, especially
if you match small-scale digitization with a well-planned
strategy for persistent access. The new generation of
scanners, with their ease of use, portability, and power,
present us with a new opportunity to expand our role
in the digital sphere.
Terence K. Huwe is president of the
Librarians Association of the University of California
and director of library and information resources at the
Institute of Industrial Relations, University of CaliforniaBerkeley.
His responsibilities include library administration, reference,
and overseeing Web services for several departments at
campuses throughout the University of California. His
e-mail address is firstname.lastname@example.org.