Kurt Molholm's Lessons Learned
by Barbara Quint
On Nov. 1, Kurt Molholm retired from his position as administrator of
the Defense Technical Information Center (DTIC). During his 44 years of public
service in the federal government, he was also very active in organizations
serving the information profession.
Molholm took over DTIC in 1985, the same year I left a long-standing position
at a leading federal contractor. The years I spent working with federal government
bureaucrats, particularly the information-oriented ones and most particularly
the information distributors, had given me a warm regard and strong respect
for them. It's always wonderful to find a fellow missionary, even one in
the process of retiring his sandals and cowl.
Q: How has the role of the federal government changed over the decades
in producing sci-tech research and in promoting its distribution?
A: Two things strike me as significant conditions that have impacted
sci-tech research. First, as Jane Bortnik Griffith stated a few years back,
is "Science as a Business." In the past, science was done first for science,
not for business profits. Science is driven by the expectation of products
more than it was in the past. Now, much of the time, basic research isn't done
until people can see its application, particularly in biotech with its race
to publication and intellectual property rights. It's always been that way
to some extent, but it wasn't at the same level 20 to 30 years ago.
The second factor is the impact of 9/11. Homeland Security and related interests
have given us almost a new Cold War impetus for research.
Q: Where do you see the federal government's role going in the future?
A: "Big Science" projects such as very-high-speed computing, networking,
and massive bandwidth. The OSTP [Office of Science and Technology Policy] report, "Research
and Development Funding in the President's FY 2004 Budget Request," stated
the following areas for funding:
The Physical Sciences
National Nanotechnology Initiative
Networking and Information Technology
K-12 Mathematics and Science Education
The federal government's fiscal year (FY2005) began on Oct. 1, but the FY2005
budget is far from complete. Only DOD and DHS have final budgets. There are
substantial increases in the defense and homeland security R&D portfolios
that will bring the total federal R&D investment to record highs. But in
the remaining non-defense, non-Homeland Security R&D portfolios, there
are big differences between the House and the Senate, with large increases
in Senate action contrasting with steep cuts in House action.
Q: Though the federal government may never reach its 1960s heyday
as a research funder across the board, are there key sectors or subjects where
it may rise again or hold on to its prominence?
A: Federal research has a narrower focus than in the '60s. There's
not as much social policy research done. I see more positive developments from
work on digital libraries and other types of research. I worry about the raison
d'être behind information system design. We're not just supposed to be
writing computer programs or building networks to move bits or bytes, but to
help people do things. Our job is to put things in minds rather than containers.
I'm now beginning to see software that recognizes that need. XML is the first
design thing I've seen that links both content and technology together. Information
transfer needs to send context. You need context to create new content. The
future lies in dealing with small digital objects and putting them together
in different ways, ways that involve the people receiving as well as those
The Net and the Web have brought about a fundamental change in human communication.
For the first time in history, technology now allows two-way mass communicationcommunication
through blogs to 10,000 of your closest friends, none of whom you've ever met
before. It's one to many and many to one. This is an exciting change, a basic
Q: How important are committees where government bureaucrats gather
to set policies?
A: Most of the "government bureaucrats" I know that work these issues
are 1) information professionals within a culture of customer support and service
imbued in them, and 2) nonpolitical. Many come from libraries, and that breeds
a different culture. When I moved to DTIC, I found an entirely different customer-oriented
culture than I'd ever seen before. It was wonderful. The common purpose was
to serve customers and change knowledge, to allow knowledge to grow.
For example, CENDI [http://cendi.dtic.mil] was originally an acronym for
Commerce, Energy, NASA, Defense Information Managers Group when it began in
1985 with just those four organizations.
Now we have 12 members, including the GPO, NARA [National Archives and Records
Administration], and the NSF [National Science Foundation]. We wanted the
top executives, people who could make things happen, rather than the lower-level
folks (who make things go). Morning sessions of our meetings, when the principals
meet, are open to ALA, SLA, the OMB [Office of Management and Budget], and
other exchanges with the outside world.
Q: How do the different information services in the federal government
work to support public service? How has that changed? Now the federal government
orders many of its tools while standing in line with private sector firms or
citizens. Is it a state of the market, rather than a state of the art, world
for the Fed and its information centers?
A: First, the federal government shouldn't control the process. Federal
organizations need lots of stuff we used to contract from industry to do our
own jobs. Put a computer system in, and with a little additional money you
can open the service to the public and reduce costs too. We can improve service
to the public. Sometimes that ability alarms firms in the information industry
that build on federal information. But at the point technology offers the government
alternatives that are better for the publicthough not for that industry
segmentthen industry must move to add different values. Entrepreneurship
means taking risks.
Second, right now we're on the cusp of a lot of software that can deliver
specific content (e.g., good enterprise-oriented systems and search engines).
Organizations must buy these systems; we don't want to develop them ourselves.
No matter what you get, however, it must be adapted to your needs. The DOD
needs all interfaces to be compatible, even though it does not want all information
Q: What protections should the government use to guarantee the historical
record? How can we build permanent public records in a digital world?
A: About truth, I can't answer; that has too much politics and opinion.
I believe facts aren't negotiable, but opinions are.
Two things are being worked on of major interest to everyone in or outside
government: open access and digital archiving. In this area again, the NLM
[National Library of Medicine] is a leader in setting a policy of permanent
public availability. At the GPO, the new public printer is looking at it too,
and GPO's relationship with NARA looks interesting. Permanent public availability
is the issue. We need systems that can deal with 40, 80, 100 years of content
in different formats, and that means finding the money to do that.
Q: What should be the role of the depository libraries? Should they
become mirror sites? Should they stake out federal sources to guard against
A: Federal organizations themselves do a pretty good job. We don't
change the record; our information professionals are apolitical. As for depository
libraries, there's a question in my mind whether they'll be needed in the future.
I have the same question about NTIS. It's good to ask questions about what
could be, how many mirror sites we would need.
As for protecting the historical record, DTIC was challenged after 9/11 to
review its documents for material that could help terrorists; we pulled 6,000
documents out of 2 million. Some material the federal government collects may
be unclassified, but still shouldn't be released publicly (e.g., privacy information).
As to the physical documents in depository libraries, DTIC is working with
some companies to improve digitizing. We need to bring the cost of digitizing
down. The period of transition is not short. No matter how high-speed the bandwidth
or [how] fast the printers, we're a long way away from not needing central
Q: Where do you see libraries and information centers going over the
next 25 years? Where do you see information professionals working in the next
A: Information professionals must recognize their primary jobs are
gathering, organizing, selecting, and assisting. Those will still be their
jobs, whatever they're doing. The information overload problem is fantastic.
The problem of information overload puts all this stuff out there for non-information
professionals until they're stunned. It's like the telephone book. When you
use it, all you want is one number and all the rest is junk.
Q: Is it critical to get Google and/or Yahoo! to take on the full-access
role for government data? How would one do that?
A: No, they're just general purpose. In government information services,
like those in CENDI, we can help our customers find what they want to help
them do their jobs. We automatically categorize material and help people search.
Better engines are coming, but individual organizations must pay for and help
build them. Our systems should recognize multiple audiences and different levels
of interest and concern with different directions and pasts. The science.gov
service sponsored by CENDI offers deep Web searching with Google-style capabilities.
Q: For the record, what do you think of the NLM/NIH move toward open
access? Is it a viable model for other government funding agencies?
A: To succeed with that model, the agency must be a dominant funder
and have an environment that can take advantage of whatever the organization
has that can contribute to publication. For example, this is not necessarily
true of the DOD. DOD contracts require documentation. About half of DOD contract
documents get publicly available both through DTIC and NTIS.
GPO Access offers another type of model: the open Web. That's a good one
too. Following that model, DTIC has approximately 150,000 documents on the
open Web on Public STINET [http://stinet.dtic.mil]. The same documents go to
NTIS, which sells them while we don't. NTIS has to recover their costs by law.
NTIS may not be able to survive with
that revenue requirement. They can't advance their services at the same level
DTIC does; we have appropriations. GPO is geared for public release, but I
think there may be problems when government documents include copyrighted excerpts.
We all must learn to live in the new world. Open access archiving is a key
Q: Which kinds of restraints on government information flow seem reasonable?
Which are unreasonable? Is there any hope of getting Congress to set a consistent
policy or at least a clearly constructed variable policy?
A: Whether you call it permanent public access or permanent open access,
there are always exceptions. For example, even FOIA [Freedom of Information
Act] has 10 exemptions, including unclassified but not publicmainly for
privacyand accessible by authorized personnel.
Sometimes people do overprotect themselves, but let me give you another scenario.
DTIC runs an independent R&D database to which private and independent
companies provide data. The data from them is lent to us, not done under contract,
and it is not publicly available. If we released it, we wouldn't get the data
anymore. Reality is pragmatic. If openness prevents giving, then we can't solve
our problems. But DOD does have a release policy that makes material available
unless someone can give a reason why not.
As for the [USA] PATRIOT Act, we need part of it, but I am concerned about
people's privacy being broached. In legislating the PATRIOT Act, people threw
lots of things in; some things need repairing, where it went too far. DTIC
was the first agency in the government to refuse to share Weblogs of usage
when pressed under FOIA.
The patenting of research out of a government contract is a very big issue.
CENDI has done a good job of putting together an FAQ on government copyright
answering all kinds of questions. I'm very proud of this; it was put together
by a DTIC staffer [http://cendi.dtic.mil/publications/04-8copyright.html].
Q: What can vendors do to work better with the government? How can
vendors help the government help the citizens? How can vendors do better and
help themselves? How can private citizens or the private sector help or interact
with such committees?
A: In Congress, staffers make things go and they like to talk to working
folks, people with problems. The general public can communicate with staffers
by e-mail or personally; so can vendors, though vendors have their lobbying
Q: And on a personal note, was it good for you? Will it be as good
for the next guy?
A: For me, it's been the greatest. The next guy will probably say
the same. You have to have a thick skin. People don't necessarily like you,
but you must prove they can respect you. You have to stay involved with organizations
around and outside the governmentfor example, working with NFAIS and
SLA. You need that so you can understand the world and the world can understand
As a general group, my colleagues in the information distribution part of
government are more interested in serving the public; the information industry
is more interested in selling product. The tension between government agencies
and vendors lies in that all vendors of information should have some altruism
built in, and I don't see that. The purpose of information is to increase knowledge
and increase profit from that increase of knowledge. If you turn that around
and make it the opposite, where the purpose of information is to make profit
first, then that narrows everything.
We've talked about concerns with the federal government and missing information,
as well we should, but government information professionals are really concerned
with the privacy issue. Show me the equivalent in the information industry
where people sell information all the time for marketing reasons. There's more
to be concerned with outside the government. People should have some concerns
with us, but we're not the bad guys, at least among the professional bureaucrats.
We are well-prepared, well-educated, work hard with the information industry
and professional societies, hold positions in all different kinds of professional
organizations. And we're dedicated.
Barbara Quint is editor of Searcher. Her e-mail address is firstname.lastname@example.org.