CENDI PRINCIPALS AND ALTERNATES MEETING
Library of Congress
Washington, DC
July 17, 2007
Minutes
Complexity and Planning
Complexity Science: What It Is and How It Applies to Organizational Transformation
Library of Congress Showcase
Responsibilities of Depository Libraries to Federal Agencies’ Material Preservation and Issues of Authentication of Digital Documents
Mr. Ryan introduced the topic by stating that Dr. Elliot Siegel of NLM had brought Complexity Science and the Plexus Institute to CENDI’s attention based on their use of this methodology at NLM. The members have previously discussed this topic as a potential approach to CENDI’s strategic planning during the annual planning meeting. However, it was determined with Plexus that an introduction to the topic would make the best use of the limited time during the planning meeting.
Complexity Science: What It Is and How It Applies to Organizational Transformation
( Lisa Kimball, Group Jazz/Plexus Institute)
Complexity is a science of change based on questions like, “How can something new, such as a new species or dynamics, emerge?” It provides a framework or a “complexity lense” through which all of science can be seen. Complexity science is a surprising convergence of a number of disciplines. It has its roots in sociobiology, math, and predictive modeling. It represents a distinct shift from the notion that there is some objective reality that you can know.
Previously in the Newtonian context, the world was linear and predictable. Scientists believed the future was knowable if you had enough data. Dissecting discrete parts would reveal everything about the whole. Phenomena could be reduced to cause and effect relationships. The role of science and technology was to predict and control this future, and our quality of life could be improved by increasingly man’s control over nature. In science, this is seen as search for the basic building blocks. In management, the metaphor is of a machine where the organization, people, and technology are no more or less than the sum of all the parts. To some extent, the computer was the worst possible thing to happen to organizations because it became possible to continue the viewpoint that if you could crunch enough data, you could know and predict.
As complexity theory developed, it discovered that this is not the best way to figure out dynamic systems because some components are not additive. The more appropriate view is that of biology rather than physics. Just as in biology or ecology, systems are somewhat simple, but they are more than the sum of the parts. In other words, they may not be complicated, but they are complex.
A model different from the Newtonian model was developed. The focus is on Complex Adaptive Systems (CAS) or Responsive Processes that are based on patterns of interactions. A CAS implies that there are patterns of interactions within them, that outcomes emerge, and that how systems behave, not how we think or expect them to behave, is important. A system implies multiple, interdependent, and connected agents. Complex implies that these systems are diverse, with many elements and a large number of connections. Adaptive implies that the system has the capacity to alter or change, making it hard to predict.
CAS have many properties. They are embedded, meaning that each CAS is made up of other CAS and is itself part of a larger CAS. These systems are diverse, having different elements which allow the CAS to change, but these elements are not interchangeable. Control is distributed and shared by many elements, rather than a single command center. In a healthy adaptive system, order and disorder coexist. The long history of our US Constitution is a good example of a document that has elements of a CAS. It is liberating. It allows basic freedom with underlying structure; intent with flexibility; order with disorder.
Because CAS are nonlinear, a small change may produce a large effect. On the other hand, a large change may have little or no impact. Ultimately, the outcomes are unpredictable. You can only consider what you can do in your sphere and within a group to make some change. While there is an underlying structure, there is also a degree of freedom. These “liberating structures” allow things to happen.
CAS outcomes emerge through a process of self-organization which creates a new whole rather than through centrally planned or directed processes. Many attribute organizational complexity to the existence of bureaucracy. However, when the French introduced bureaucracy to government, it was not meant to be so constraining. What developed is that current systems are more about putting checks on what happened yesterday, resulting in layer upon layer of what we now call bureaucracy. A synergy between the formal and informal structures makes for the best organizations.
These attributes of CAS are interdependent and nested. Ultimately, it is about the network. A hierarchy is only one kind of network. It is important to think about the organization as a network, not as an organization chart. How do resources travel? What is the network hub? Networks are also perpetually self organizing. Order occurs from the grass roots up.
An organization can be viewed based on the stages of network development. Networks have different stages of development. Stage 1 is a scattered cluster of unconnected individuals. Stage 2 shows the beginnings of more organization with a hub and spoke, with something more central drawing the individuals together, but the connections among the individuals are still rather loose, going only through the hub. Stage 3 is a multihub environment with a series of smaller networks. Stage 4 is the most complex with a core and periphery or a smart network. Networks get more powerful as you work through these development stages. So the question becomes, what can you do on purpose to improve them? Or, in looking at a network of interest like CENDI, what stage is appropriate for the goals you want to accomplish for the network?
The group discussed the application of complexity to CENDI as an organization. At present, CENDI, at the least, is a Hub and Spoke-type network where people learn from one another but most of the actions occur through individual members’ centralized organizations. The comment was made that perhaps this is why CENDI is excellent at networking, but until Science.gov came into being, it was not very effective at an on-going, CENDI-based operational project.
The next stage is a network made up of smaller sub-clusters where there are close connections due to common interests. It is possible to draw a circle around all the small networks to begin to identify the core and the periphery. Plexus has software to help create the network maps. These maps can identify weak spots and measure the closeness of individuals within the network. The network map gives people a way to talk about their connections and how to improve the interconnectedness.
Ralph Stacy has developed a matrix approach to address the degree to which the group is close to certainty. If you are already close to agreement and certainty, this is a much simpler approach. As you move farther away, you get more complicated but there are still some systems and processes in place.
The traditional view of diversity in organizations would focus on responding to the environment and addressing a problem to be solved. The complexity science view is all about maintaining complex relationships, learning in real time, and becoming a strategic resource. There is a “gray zone of complexity” between a simple solution (close to agreement and close to certainty) and anarchy (far from certainty and agreement). This is where creativity can happen.
What is the appropriate strategic planning when you are trying to manage the unknowable? First it is important to accept an unforeseeable future. You must accept nonlinear outcomes and the inability to predict. The key is to strike a balance between clockware, a very regimented approach to organization, and swarmware, where nothing is organized. In this context, the complexity framework focuses on doing – planning, liberating structures, timing, embedded work processes, many small actions, and a focus on the whole system.
The value is not in the plan but in the activity of planning itself. It is important to look at a portfolio of interests and to address issues of timing. The analysis of the situation must be constantly repeated. There is power in identifying the right number of simple rules and specifications. Innovation is based on a lot of messy stuff. The focus should be on “failing fast”. The more traditional budget process is offset by the entrepreneurial investment process. That is bid a lot and hope for a few big winners. Venture capitalists pick 17 and hope to hit one rather than pick one and put all their eggs in that one basket. Other approaches to innovation include “serious play” and prototyping that involves external models. One example of this would be the Gore Reinvention Laboratory.
Social diffusion and adoption are necessary. Ultimately, the changes must be adopted by the organization, crossing the chasm between the early market visionaries and technology enthusiasts and the pragmatists of the mainstream market. How do you cross this chasm in a good way? Scalability is an issue; the mistake is often doing the same thing but bigger.
Complexity scientists understand that knowledge doesn’t change behavior. More education doesn’t necessarily change behavior. If it did, we wouldn’t have smokers, etc. While peer-to-peer communication and sharing best practices is important, the focus must be on practice rather than knowledge. Acting in new ways is more important than thinking about them. Complexity also looks at why people don’t act in the way they know they should. The cause often leads to different paths of action.
It is also important to manage the lifecycle of the system and to make people more aware of processes. The life cycle must be managed not on the S curve model of birth, growth and maturity, but as an infinity cycle where the system is constantly being born, renewed, reaching maturity, being destroyed, and reborn.
Discussion
The group discussed how complexity science and network theory might relate to CENDI and its planning process. CENDI has grown and changed over time. It is important to look at the current strengths and weaknesses. We need to look at strategic positioning. What is CENDI uniquely positioned to do at this point? What is its network and where might we want to go in terms of strengthening or adding connections and why? For example, how do we interact with the policy or research community? Do we need to strengthen that “network?” As CENDI has grown, it has increasing success in creating communities of twos and threes. Often, the connections between agencies that occur at the periphery go unrecognized and undocumented. It is also important to realize that not all nodes are equal at all times. There may also be strength in using alumnae and introducing younger people to the network. We should define what our core is and what our periphery is and what we want it to be. The difference leads to paths for action.
Library of Congress Showcase (Roberta Shaffer)
At the time it was built, the Jefferson Building of the Library of Congress was the building that envisioned America and what the country would become. A major question for the Library now is how to continue its mission into the 21st Century. The Library is expanding its constituency beyond the physical walls of the Library, and seeking, through agencies like CENDI, to expand its collaborations. The Library staff highlighted a few examples out of the many projects the Library has undertaken.
Introduction to the Science, Technology and Business Division – William Sittig
The inspiration for science in the Library of Congress goes back to the donation of Thomas Jefferson’s Library after the original library burned. Most of Jefferson’s works are still in the Library, and they will eventually be put on public display. Jefferson’s science collection was enhanced when, in 1866, Congress transferred the Smithsonian collection to the Library of Congress. The actual Science and Technology Division dates back to 1949. A Business Division was added about 10 years ago. The Division covers all areas of science except those covered by the other national libraries.
The Library has several interesting science collections. The Library has manuscripts and rare books collections from many important scientists and first editions of the 100 most important volumes in science. The Library is very interested in making its collection available through Science.gov. One of the additions to the web catalog will be a web site called “Everyday Mysteries,” which answers interesting questions such as “why is the sky blue?’. This is the second most accessed site within the Library of Congress web domain.
Technical Reports and Standards – Lawrence Marcus
The S&T collection of this library is primarily for researchers. There are databases and electronic resources available in the reading rooms, and access to a collection of approximately 100,000 standards and more than three million technical reports dating back to the post-WWII era. For example, the Library has the documents from the Office of Science Research and Development (OSRD), which was headed by Vannevar Bush. There are key documents related to anthrax, pesticides, etc., that are still valuable today. The collection includes conference papers, NATO documents, and full runs of AGARD, AIAA, and other organizations. A specialized guide to aeronautics is being developed. They also have Japanese and Chinese translation of technical reports on the Japanese medical experiments after WWII. Technical reports are available in both paper and CD.
Technical reports are cataloged only at the collection level; DTIC and NTIS catalog records are used at the item level. The Library is interested in complimentary collections from other agencies. For example, NARA has the OSRD administration documents while LoC has the technical reports. Mr. Marcus encouraged agencies to contribute their technical reports to the Library.
The Library hopes to encourage digitization of the technical reports. Google was interested in digitizing the Library’s collection, but it could not be done through a mass digitization project because much of the paper is too fragile and the documents are of odd sizes.
Databases and Electronic Resources are available through the Library. Some of these databases are subscription-based, such as the CSA Technology Research Database, ProQuest, and Jayne’s. Some backfile collections have been purchased but these are in the minority. They couldn’t cancel print subscriptions to allocate money for the electronic versions since the print comes to the library for free with copyright deposit. The licenses for these resources generally limit use to on-site at the Library; they are hopeful that more online access can be negotiated in the future.
E-Deposit – Baha Akpinar
The LoC’s E-Deposit Working Group is working with service content providers, publishers, and archival services to enable publishers to submit electronic works for copyright deposit. This group crosses LoC divisions and specialties and involves software developers and IT professionals as well as librarians and archivists. Copyright deposit has been a key avenue of input to the library’s collection in the print environment. Ninty-nine percent of the current e-deposits are in the public domain; the problem is they are copyrighted. The working group is starting with those materials without print counterparts, which are at higher risk for being lost.
E-deposit involves content transfer capabilities, the structure and development of a preservation repository, and a set of policies. Some aspects of the high level design have been developed, including a pilot repository. A major challenge has been dealing with the uneven adoption of standards by publishers.
The e-Deposit project builds on the CORDS Project, a Copyright Office project where 20 publishers are depositing electronically. However, the solution for each deposit stream is handcrafted for each publisher. The working group wants to identify a general solution that builds a consistent repository that interfaces with the ILS and other library services.
An alpha test was conducted in January/February 2007 with content from three publishers. They took the content, parsed and mapped it to a canonical view. The curator can view the results and refine the process. On the policy side, they have developed a best edition statement. Several policy roundtables are scheduled this fall with publishers and stakeholders about what access can be provided by the library. The system is expected to be in production by the fall of 2008 with STI e-journals.
Preservation Activities – Dianne van der Reyden
In its mission to preserve a universal collection of knowledge and creativity for future generations, the Library’s Preservation Directorate deals with a variety of scientific and technical disciplines from materials science to computer science to chemistry. For example, they deal with original chemical composition using techniques such as ultraviolet and infrared analysis. Through its own Research and Technology Division and with special collaborations with others, such as the Institute for Museum and Library Services, the Society of American Archivists, NARA, NIST, Lawrence Berkeley National Laboratory, and academic institutions, the Library performs a variety of research into how to preserve and conserve the various materials received by the Library.
They have worked with NIST on the reliability of various media and on the fade rates of various inert gases, with DNA experts at Harvard in setting up clean rooms, and on other projects such as mass de-acidification and development of a fingerprint to diagnose sick versus well magnetic tapes. Through an NSF grant, the University of Washington is developing computer simulations to enhance the skills of conservators. The Directorate has had over 100 fellows and interns and has conducted workshops supported by the Mellon Foundation.
In addition, the Preservation Reformatting Division has created the National Digital Newspaper Collection with support from the National Endowment for the Humanities (NEH) and the Open Source Initiative (OSI). The goal is to identify good practices that can be implemented at a variety of libraries. The Binding and Collection Care Division is responsible for the ongoing care and preservation of the regular library collection, including the development of innovative housings for new and mixed media items. In addition to preservation of digital assets and ongoing research, the Directorate will continue its focus on emergency preparedness in the 2008-2013 Strategic Plan.
Strategic Planning in Library Services – Deanna Marcum
The Library Services Division is working on a strategic plan for 2008-2013 to address where they need to be in five years. The major strategic initiatives include bringing in collections and preserving them, developing collections in a digital era, connecting users to content, improving cultural understanding, and increasing the LoC’s leadership within the library community, now that every library can more easily contribute to the community. The Library must also manage for results, determine performance measures, and convey them to employees and stakeholders. In total, there are 42 strategies and 38 working groups of LoC staff tied to budget priorities.
As the resources of many libraries for preservation activities are decreasing, there is a role for LoC. They have spent over two million dollars on a testing laboratory. They are also working on issues of bibliographic control. These discussions have been very heated and have raised profound questions. It involves moving beyond the catalog card and beyond a central environment. There are materials available from the LoC web site from these discussions of standards, structures, use, bibliographic control, and economics.
The Library is thinking about bibliographic control in the 21st century. They are raising profound questions since we can’t keep making everything look like a catalog card. They are looking at standards and structures for bibliographic control with a heavy emphasis on economics. A working group on Bibliographic Control has met in three geographic regions. A report is due at the end of August for public comment; the final report to be due in November. The results will be used for the Library’s strategic thinking in this area.
A two million dollar grant from the Sloan Foundation has been received for digitizing the general collections. In the past, LoC focused its digitization efforts on older collections in American History. However, they are determined to take collections to the users. With 135 million items in the LoC collection, they must think through strategies for digitizing. The current efforts will deal with a broad scope, and will involve issues of selection and workflow for different kinds of materials. LoC will start with the public domain materials first.
On the whole, there is bi-partisan support for the LoC. The biggest challenge is how much libraries should collect, because Congress often has the misperception that it is all on the web.
Special Interest Topic: Responsibilities of Depository Libraries to Federal Agencies’ Material Preservation and Issues of Authentication of Digital Documents
(Prue Adler, Association of Research Libraries, and Peter Young, National Agricultural Library)
Ms. Adler introduced the topic from the Federal Depository Library Program (FDLP) perspective, but she indicated that there are broader issues. Fifty-two federal depository libraries (FDL) are charged with holding government documents in perpetuity. (ARL is involved: between 23 of the 52 FDL members are ARL members). However, users want electronic content and different services than those that can be provided with paper. The digitization that is being done with Google and others is being done to access rather than preservation quality standards. In addition, there are issues about authenticity of official government documents.
These projects raise a lot of questions including how to authenticate government documents in electronic form, what level and how to apply metadata and cataloging, and how long FDLs need to keep the print if there is access to the electronic. Is there a way to cut through some of the old policies? Ms. Adler believes that some of these policy issues are better addressed by the department or issuing organization. For those agencies willing to tackle these questions, the academic and state libraries would provide many partners.
GPO is starting to authenticate the born-digital materials, but the bulk signing is not yet available. However, the technology is somewhat separate from the policy. What is the official version? In print, the provenance is clearer, especially for agencies who receive the materials through official program channels with RDP pages and signatures.
Ms. Adler asked that CENDI consider exploring these issues jointly and within the individual agencies. The Policy Working Group asked if there is interest in doing a pilot or having further discussions. A number of agencies expressed interest in further discussions including the NAL (Peter Young), NASA (JoAnne Calhoun), GPO (Rick Davis) and NTIS (Ellen Herbst).
