As Howard Bloom described the speakers as partners in creating new knowledge, we are partners in enabling the creation of new knowledge. The collections in your institutions are passive, as is the content in all of RLG Cultural Materials. While it takes interaction to create new knowledge, existing knowledge, information, and resources are often springboards for new discoveries.
No one knows more than you about the symbiotic relationship between library collections and new scholarship and about the importance of primary sources to researchers. Early results from a Mellon-funded study seeking to learn where students and faculty go for information particularly identified image resources as desired and as underserved. They are very important to course work and to research; yet, they were described as having serious access and discovery issues.
Many of you have already contributed to the RLG Cultural Materials effort, to bring together the special collections of premier libraries, archives, and museums into a single resource in a way that will make possible many exciting discoveries, connections, and new avenues of research. Ian Foster's definition of the science grid focuses on high-performance mechanisms. The emphasis of the culture grid is on access to resources. What these grids have in common is an opportunity to stimulate significant change in research by sharing resources on an unprecedented scale.
In April 2002, there are 47 RLG members participating in the Cultural Materials Alliance. Based on what alliance participants have told us will be available for contribution, the number of contributed collections will increase from 50 to over 150 in the next year. Of course, the number of alliance participants is also likely to increase, which would considerably boost the projected growth of the resource.
Right now we are busy preparing the next 11 collections from 11 different contributors, but we are always looking for what is ahead in the queue. The number of items could increase from 100,000 to nearly 900,000 in that same year. An item could be a photo, a book, or an hour of video. While a picture may be worth a thousand words, these items are more different than apples and oranges. They are more like grapes and truckloads of watermelons.
RLG Cultural Materials became available to researchers this January. At that time there were just over 50,000 items from 36 collections from 18 contributors. Today there are over 100,000 items in the resource. The collections we are currently working on will add another 76,000, bringing us closer to 200,000.
Today we will look at three different ways that RLG Cultural Materials can act as a catalyst in enabling the creation of new knowledge, and ways in which we can all actively play a part. We'll do this in three sections: Surprising Connections, Growing the Resource, and Optimizing Discovery.
The nature of the collections in RLG Cultural Materials is now somewhatrandom. To begin building the resource we included collections that hadalready been digitized. You all have very good local reasons for selectingcollections to digitize. Some were digitized because they were easy, somebecause they had preservation or access requirements, some because theyfit the theme of a collaborative project or because of an event or an exhibit.
When these collections are brought together into a union database, the selection criteria do not necessarily result in a cohesive whole. For instance, if you combine celestial maps from the Linda Hall Library, rare Shakespeare folios from the University of Pennsylvania, calotype photographs from the University of Glasgow, Russian children's books from the International Institute of Social History, medical history photos from the University of Edinburgh, the Farber gravestones from the American Antiquarian Society, rare books from the Smithsonian, illuminations from the Book of Hours from the Huntington Library, the last letter of Mary Queen of Scots from the National Library of Scotland, and the Wilfrid Owen Multimedia Archive from Oxford, what do you end up with? Certainly a lot of wonderful artifacts, but finding the convergence among the "adhockery" is difficult. Fifty narrow collections do not create a resource with depth and breadth. Five hundred collections start to be really interesting.
Even in these early days the modest resource reveals some surprising strengths, and interesting connections abound. This is the service that RLG staff demonstrate most eagerly. Invariably the visitor brings an interest and some knowledge to an avenue of inquiry that often turns up the delightfully unexpected.
Perhaps the most significant benefit of RLG Cultural Materials is the bringing together of disparate resources from repositories in various parts of the world. Some of these would never be thought of individually to shed light on a given topic, but discovering their hidden content serendipitously is a pleasure.
With the pooled collections, as varied as they are at this stage, to make the most of the resource the burden is on the scholar to fire his own synapses and make informed inquiries. The researcher may also rely on the traditional mediator role of librarians, archivists, and curators to identify different terminology to search for things or to relate content to other resources to interpret what he finds.
Eric Ketelaar employs Dutch anthropologist Geert Hofstede's phrase "the software of the mind," saying that in archival usage the socially and culturally determined software of the mind also plays a role. In his paper "Research in and on Archives," presented at the National Scholarly Communications Forum Round Table in Canberra, November 1999, he says:
Now let's focus on finding synergy amongst 45 very diverse collections.
It is interesting to follow a rather random entry point into the resource. We type in a keyword search on the word "blacksmiths," and take a look at the diverse collections that have blacksmith-related items, while getting a glimpse of blacksmiths and their workplaces over time.
First we take a look at a dispatch card from the 1800s for a clipper ship. This shows three men, a blacksmith, a sailor, and a soldier on a ship deck. Blacksmiths loomed large at sea, since ships require all manner of pulleys, cleats, brackets, chains, and repairs. This is from the Robert B. Honeyman collection in the Bancroft Library at Berkeley.
Next is a daguerreotype from Berkeley's Cased Image collection, showing Sacramento in 1852. The blacksmith shop is a little bit off the beaten path, but next to the Tin and Sheet Iron Ware shop.
From the National Library of Wales' Framed Works of Art collection, we have an 1866 Whistler etching titled "The Forge."
From the Chicago Historical Society's Chicago Daily News photo archive, we see a traditional view of a blacksmith toiling over his anvil next to a brick oven in 1906.
We catch a blacksmith in the act of shoeing a mule. This is from Indiana University's Frank M. Hohenberger photo collection. It was taken in North Carolina in 1928.
Finally, a poster from the Works Progress Administration collection contributed by the Library of Congress. This 1939 poster aims to encourage laborers to work for America.
That is just a few of 49-odd items about blacksmiths. A researcher wishing to illustrate a paper on the evolution of blacksmithing would bring all sorts of other angles to bear on the topic.
Now we will do a search on "mining machinery" and pull up a 1925 image from the pictorial collections of the National Library of Australia.
Blacksmiths shaped much of the new world figuratively and literally. Without tools, man conforms to the land. With tools, man can make the land and its resources conform to his wishes. This new power led to more horses and wagons, more roads, and more communities. Blacksmiths were critical to this development—and to the westward movement.
This got me thinking about geography, so I type in a search for "iron and maps." Of course, blacksmiths and forges were everywhere there were horses and construction, but the larger operations were near iron, coal, and waterways which later attracted railroads.
From the Library of Congress map collection we see a map of Ironwood, Michigan, showing a bird's-eye view in 1886. The key indicates the location of the blacksmith shop. To the north is the mining operation. Not far to the east is the Montreal River, and to the south of the blacksmith shop is the railroad.
On a larger scale we see a map from 1881 from the Ohio Iron Company. This map shows the railroad network that centered around the mine and coal field as well as its relation to the markets in the north and west.
To pursue another angle on this story, beginning in the late 1790s the cotton gin and other new developments improved production rates and quality, reduced costs, and increased exporting, all resulting in more plows forged, more wagons, more parts and more repairs—thus more work for the blacksmiths, including railroad construction. Railroad construction depended on blacksmiths for all the tools and spikes and so forth.
A search on "railroad construction" turns up this 1936 photo, from Indiana's Hohenberger collection, of work on a railroad near Pendleton, Indiana.
We also have a series from the Pennsylvania Bridges collection from Penn State University.
Take a look at the Warren truss viaduct. This was built in 1906 for Pittsburgh Railways.
This collection also includes technical drawings. Here is a detail from an engineering drawing which we can zoom in on at great levels of detail to see diagrams of stresses and sections of the railroad bridge.
All these developments spurred new commerce. While blacksmiths had a large role in building railroads, the railroads had an even larger role in the demise of blacksmiths. By the late 1800s the railroads had linked the country, and hardware was being mass-produced at plants and sent by rail to be sold in hardware stores.
A search on the word "hardware," pulls up a photo from Duke University's William Gedney collection showing a hardware store window at night.
Image from RLG Cultural Materials. Not for reuse or redistribution.
With the coming of the industrial age most blacksmiths were relegated to the maintenance departments in the plants that had replaced them. Doing a search on "factory," our result includes this image of the Union Iron Works in San Francisco in the 1880s.
Image from RLG Cultural Materials. Not for reuse or redistribution.
Emerging mass-production techniques began to replace much of the blacksmith's work. Hand-crafted iron was largely superseded by manufactured cast steel. A search on the word "automobile" finds this image from the University of Minnesota Performing Arts Archives. It is a turn-of-the-century ad curtain showing a commercial street with an automobile. Affordable automobiles and tractors soon made horses and wagons nearly obsolete.
Image from RLG Cultural Materials. Not for reuse or redistribution.
This 1945 pamphlet from the London School of Economics is entitled "A Car Traveling People." It is described as being a scene-by-scene picture of how drastically the first 40 years of automobiles had altered American life. For instance, when a USDA investigator asked a farm wife why her family owned a car when it didn't own a bathtub, she said, "Why, you can't go to town in a bathtub!"
Image from RLG Cultural Materials. Not for reuse or redistribution.
Between 1910 and 1930 most practicing blacksmiths shifted their attention to architectural "ironwork." Many of our most treasured works in iron were made during this time. A search on ironwork finds this contemporary painting from the National Library of Wales.
Image from RLG Cultural Materials. Not for reuse or redistribution.
A search on "metalwork" reveals a Federal Art Project poster from the Library of Congress promoting the New York City Craft School, where adults could learn pottery, modeling, weaving, painting, drawing, wood carving, sewing, needlework, photography—and metal work. Today, the blacksmith survives, but mainly as a specialty artisan. The quality of the work is more impressive now than it has ever been. But, rather than keeping the wheels of progress turning, today's workmanship is primarily sold to collectors.
Image from RLG Cultural Materials. Not for reuse or redistribution.
I hope you will agree that, while we don't have what you would think of as "go-to collections" on blacksmithing, an enterprising student could pull together an interesting array of materials to illustrate an essay on the topic.
The scholar creates his own circumstances, and different scholars could take different paths through the subject matter—and through the content of the resource.
To encourage the making of novel connections between disparate collections, we need to think about how to empower the researcher. How can we increase awareness about the availability of new resources that the researchers may not discover on their own? How can we best instruct or assist them in using unfamiliar resources? How can we build on the traditional roles of librarians, archivists, and curators to mediate between researchers and the variety of resources available to them? Where are the opportunities to encourage faculty to take full advantage of new modes of research?
Growing the Resource
RLG has created a content-manager database that allows Web input by staff at alliance institutions. All of the proposed collections are described and information is provided including date ranges, subject headings, geographic coverage, original format, as well as the technical information needed to harvest the files and descriptive records. The contents of the database can be sorted in many ways to create content profiles that reveal strengths and gaps in the overall content of RLG Cultural Materials. A content advisory group helps us to identify areas in which to encourage development. RLG will propose identified areas to granting organizations.
Use statistics and content profiling form an adaptive feedback loop that can have an impact on collection development. The statistics we gather can give you feedback about what types of content researchers seek so that you can leverage your collections and your digitization resources with what is actually wanted. The feedback can provide input into your digitization decisions, to be balanced with local motivations, and perhaps influence your priorities. We expect that the feedback will also be of interest to foundations being asked for digitizing funds, because they want to support digitization of content that is sorely needed. You can use the identified areas of need to support your own proposals for digitizing projects.
If the blacksmithing demonstration showed an unexpected strength, this next demonstration identifies a disturbing gap in a search for East Asian material.
We start with a keyword search on the word "Japan." We get over 7,000 results. We can take a look at the "what, where, who" view of the interface.
Image from RLG Cultural Materials. Not for reuse or redistribution.
This allows us to refine our results. These lists show how the items were described. The "what" column offers work type, subject headings, and collection names. The "where" column offers locations in which the items were created or locations depicted in the items. The "who" column offers people or organizations that created the items, who are depicted in the items, or the institutions that contributed the materials.
We select "contributors" and take a look at the variety of sources for these materials which contain the word "Japan." We can see that the vast majority are from Berkeley, so we ask to see the collections that are presented in this result set. Scrolling down the list, we can see that the majority are from the War Relocation Authority photographs of Japanese-American Evacuation and Resettlement during World War II.
Image from RLG Cultural Materials. Not for reuse or redistribution.
We take a look at one of these series. This is a series from Minidoka in Idaho in 1942-43. In this thumbnail view we can get a sense of the materials in the collection.
Image from RLG Cultural Materials. Not for reuse or redistribution.
One of them is captioned "Reading a Bedtime Story"; in others they are cooking, learning about auto mechanics, having a haircut, observing something in a microscope, taking a drawing class. This is a very idealistic view that the War Relocation Authority wanted to project. It of course clashes significantly with the harsh realities that were described by the internees.
We return to the Berkeley collections and notice that there are a couple from international exposition collections. We will choose the one from the California Mid-Winter International Exposition in 1894. Here we have a western view of a Japanese tea garden....
Image from RLG Cultural Materials. Not for reuse or redistribution.
A search on the word "Korea" turns up just six things, four maps from the Library of Congress and two items from the War Relocation Authority.
A search on "China" results in over 200 records; in this case 140 or so are from International Institute of Social History. In the first instance we see a 1953 propaganda poster urging the Chinese people to "Elect good people to do good things."
Image from RLG Cultural Materials. Not for reuse or redistribution.
There is another one entitled "Carry out family planning. Implement the basic national policy." Here the one-child policy is represented as the road to a modern, western-looking world. Even the woman and the child themselves have un-Chinese features, or so the caption says.
Image from RLG Cultural Materials. Not for reuse or redistribution.
Propaganda posters give us another skewed view of Asian culture. Let's look at other types of materials.
We can return to our "Japan" search by using the history feature in the interface. This time in the "who, what, where" view we limit the work types to "text and documents." We find a piece of sheet music from Duke University entitled "My Yokohama Girl." This is from 1917.
Image from RLG Cultural Materials. Not for reuse or redistribution.
If we look inside the sheet music, we can get to the actual notes and lyrics. Enjoy these lovely lyrics: "She just murmured 'hicky-hoi' and closed her Japanese fan. He just answered 'Ship Ahoy' and grasped her little brown hand...".
The subject headings in the descriptive record for this work include see "Legacies of Racism and Discrimination—Asian." I think we have seen enough in that vein.
Image from RLG Cultural Materials. Not for reuse or redistribution.
Let's try one more search, this time on the keyword "Asian." Here we find a woodcut from the Bayerische Staatsbibliothek offering a German depiction of the Asian King Melchior, one of the three Magi.
Image from RLG Cultural Materials. Not for reuse or redistribution.
And we have a portrait of an Asian boy—done by an Australian artist and contributed by the National Library of Australia.
Image from RLG Cultural Materials. Not for reuse or redistribution.
Finally, this picture is from the National History Museum in London, the McGillivray collection. It shows an Asian species of black-hooded oriole. This is perhaps the most authentic Asian artifact we have found, although it is painted by a Scottish ornithologist.
Image from RLG Cultural Materials. Not for reuse or redistribution.
You see that, although there is a fair amount of East Asian content, the resource desperately needs genuine Asian collections and hence a targeted digitization initiative.
A group of eight representatives from RLG member institutions established selection criteria for choosing among proposed Chinese, Japanese, and Korean collections. They then selected 11 collections, held by RLG members on three continents, containing about 35,000 images. The collections represent unique primary source materials that document different aspects of East Asia in the 17th to early 20th century. They illustrate the transition from the Imperial period to the start of the modern period and include digital surrogates of maps, photographs, pictures, manuscripts, diaries, and rare early published works. A funding agency is currently considering a proposal to defray the costs of the digitization of these collections.
We have grant funding to support future meetings of advisors to further identify areas to target and to make selections among proposed collections. Using the various feedback mechanisms as a tool, we can work together to make this resource even better.
Some of the ideas we are currently working on for expanded coverage include
- Voyages of discovery and exploration—we are currently seeking funding, after which we will turn to RLG members to nominate collections.
- Global migration—a proposal is being considered to fund digitization.
- World's fairs—this is an idea in development. These collections reflect bright moments of time and place when current industry and culture are shown off and during which we project into the future.
- Labor history—the RLG membership certainly has stellar collections in this area.
- Performing arts and indigenous cultures are among other fertile areas.
We hope maybe the next idea will come from you, and we encourage you to tell us about possibilities where you have collections that you know would fit nicely with collections from other institutions.
Optimizing Discovery
Features in the interface take advantage of intellectual links in the data to take you places where you might not have thought of going. The infrastructure and tools provided can increase the facility with which items can be found and related to others. Some of the data currently being contributed is fairly thin, but other records are rich with words and concepts. Rich records can act as a springboard to other materials. An innovative data model was created to take advantage of the variety of materials and descriptions. That model is evolving as we are gaining more familiarity with the contributions.
Since libraries, archives, and museums use different descriptive approaches, the system needs to be tolerant of different data standards and differing controlled vocabularies (or a lack of controlled vocabularies). We are investigating ways to have the interface mediate some of those differences. We rely on you and your users to influence the next functionality enhancements.
In the next demonstration, the interface and its functionality create circumstances for following trails through the content using links inherent in the data.
Let's suppose a theater class was looking for period costume ideas to outfit the leads in a production set in the mid-1800s. We will do a search for "husband and wife." The first items in our date-sorted results are prior to our target date, but it's hard to resist taking a peek. This one, from Berkeley, is a marriage contract from 92 BC. It describes a dowry of two talents and 4,000 drachmas in exchange for fidelity, which is defined in great detail.
Image from RLG Cultural Materials. Not for reuse or redistribution.
The next item is from the Smithsonian's Material Culture collection, and it is called "The Bottle. It is brought out for the first time: the husband induces his wife 'just to take a drop.'" This is one of eight drawings from 1854 showing the degenerative effects on a family due to the evils of drinking.
Image from RLG Cultural Materials. Not for reuse or redistribution.
An 1850 photo from the Berkeley Cased Image collection has among its subject headings, "Couples."
Image from RLG Cultural Materials. Not for reuse or redistribution.
Following that link in our interface, we are presented with a result set that contains another piece of sheet music from Duke University. This one is "All the World will be Jealous of Me."
Image from RLG Cultural Materials. Not for reuse or redistribution.
Looking at this image, we start to narrow our focus to women's clothing. From among the subject headings on this record, clothing and dress and costume are obvious choices, but we could also see what we get with the general topic "women," as any image of women is likely to show clothing. (You can tell from the statistics that we keep that some people are looking for the ones that don't.)
This time from the "what, where, who" view we choose photographs from the work-type category, and we see that this is a fairly good way to view women's costume.
Image from RLG Cultural Materials. Not for reuse or redistribution.
Our attention is drawn to a photo of women wearing somewhat peculiar hats, from the Chicago Daily News Photo Archive, and it is captioned, "Five women wearing the same style hat during the 1904 stockyard strike."
Image from RLG Cultural Materials. Not for reuse or redistribution.
There must be some pictures of more typical hats. Let's follow another subject heading for hats. Here we find more sheet music from Duke University, this piece for a tango and featuring a fine specimen of a hat.
Image from RLG Cultural Materials. Not for reuse or redistribution.
It makes me want to pursue the topic of dancing for costume ideas. Following the "dancing" link, we find an item from the Library of Congress Ballroom Dance Manual Collection. It is entitled "Wehman's Complete Dancing Master and Call Book: containing a full and complete description of all the modern dances."
Image from RLG Cultural Materials. Not for reuse or redistribution.
It has a related view to a waltz cotillion video, which was performed by members of the Jonquil Street Foundation and the Library of Congress Centennial Cotillion Brass Band in the Great Hall of the Library of Congress.
Image from RLG Cultural Materials.
Not for reuse or redistribution.
You have seen our first efforts at interface development to help optimize research in RLG Cultural Materials. We are planning to do much more to facilitate use of the collections.
Currently we are working on detailed statistics such as zero result searches and which collections are most often accessed. We are working on several enhancements to searching functionality to allow for selecting more than one item from a "what, when, where" view, searching within a result set, and getting more value from complex subjects—for instance, treating geographic subdivisions as place names. We are implementing METS (the Metadata Encoding and Transmission Standard) to facilitate page turning and viewing of related files. And we are developing a means to bookmark and to return to a selected item from outside the system.
In the near future we will be working on improved navigation between items and their collections, downloading files to allow local use, assisted searching to take advantage of controlled vocabularies even when the data does not, better surrogate viewing and sorting, and the ability to export to courseware. We know researchers and faculty will be using RLG Cultural Materials along with other resources, and they will want to bring all the things they find into a familiar software environment. No one is closer to the users of the resource than your staff is. We hope you will let us know what enhancements are most needed.
Knowledge creation happens best in an environment where content and tools are broad and rich. You as contributors help to develop and enrich content. You as advisors help to develop and enrich tools. You are key to creating this environment for facilitating knowledge creation.
To characterize this collaboration, I would like to invoke Howard Bloom's microbial team from The Global Brain:
[They] didn't just recycle ancient programs, [but] through data pooling, experimentation, and tests of novel strategies, the bacteria managed to refashion themselves in radically new ways. This was not traditional random mutation at work. This was driven, inspired conception.
From the Discussion that Followed
Discussion delved into the use of RLG Cultural Materials so far and explored how it could be broadened, how contributions could be encouraged, even how users might be directly involved in supplying descriptive data or targeting desired content.
Direct User Feedback
Moderator: I am especially interested in how you can follow in your use pattern what types of searches and combinations are made by your end users. Is there a possibility that, once a user has made a combination, maybe also with materials which are not yet in the database, that user can then present his or her own collection for inclusion in the cultural resources? Is it all done by the conservatories and by the institutions or do you have a sort of sharing with the user and do you invite the user to bring his or her own collection?
Ricky Erway: We are looking at ways to include all sorts of different types of material from all sorts of different sources. A good place to start, though, is with the collections of the people in this room.
Moderator: I checked RLG Cultural Materials and found some photographs, I think from Chicago, and a caption was not right. Can I send back directly to that collection, "This is not the Dutch church in Moscow, but it is the Dutch church in St. Petersburg?"
Erway: I am sure they would want to know that. It is a good idea to have a direct link. So far contributors have been very happy to hear when there are issues.
Moderator: In fact, it is not my own idea, but the Municipal Archive at The Hague has mounted on the Internet their collection of photographs. On every screen there is the possibility for the user to enter his or her comments on the picture, and they are used to enhance the descriptive data in the collection.
Erway: We get mixed reactions on this, so I would be interested to hear what some of the people here think—whether end users should be able to add data to the descriptions that are in a system like this.
Moderator: As long as you indicate that this was a contribution by a user in the year 2002.
Participant 1: If there is erroneous information, that is one thing.
Moderator: As an archivist I have always wondered: The people who know best about the content of the material we are providing are the users, but they keep that knowledge to themselves and there is virtually no possibility to share that knowledge with other users.
Participant 2: One area that seems to me to have the most possibility for user-contributed description is in the problem of providing collections that are designed and described for a certain disciplinary context. If we want to try to get use outside that context, we are left guessing what that new context is going to be and how to describe it for them.
If that is a goal, and it seems like it could be interesting to experiment with user-contributed descriptions as the users would be bringing their own contexts, it seems to me that it would be most useful to keep those types of descriptions somewhat separate from the more specifically edited or focused audience type of description.
Moderator: I agree, but the first step of recontextualization has taken place already with the description by the librarian or by the archivist. The artifact or the document has been created in a certain context and then it is moved from that context virtually or physically into the context of the library, and there we add our archival or library description. I agree that you should keep that separate and clearly marked that this description was at a given moment in a given time and even by a given archivist, but ask people to share their knowledge thereby creating new knowledge.
Participant 3: I would not want them to feel that we would not change our core record, but the mechanism you describe where there was possibility to interactively comment and then someone would vet those comments. The only drawback is what type of workload that becomes.
Participant 4: Are there two issues, though? There is the first question of correctness. You mentioned earlier that something you found was incorrect. In that case you would inform the institution providing the image that the information was incorrect. When it gets to comment, you would end up presumably with a formal description, followed by a discussion list to which people could contribute things. That might be a way of doing it.
Moderator: Wouldn't that be the ultimate and wonderful interactive library, a combination of the concept of a discussion list with your materials and with your descriptions? The separation serves keeping the authenticity and the integrity of the object plus a description, but we should be aware that even our description is a subjective interpretation or representation of reality.
Participant 5: The case I am thinking of is a collection of images from the Picture Post. We found one of these images, and the whole description is about James Callaghan, later British prime minister. He is standing inside a Henry Moore sculpture which is not mentioned in the Picture Post description of this photograph at all. I think that would be the thing that would be most interesting to many people.
Participant 6: I was wondering about that idea of getting people to contribute descriptions. People are using the resources to create knowledge, not just to give comments on descriptions. Would there be a possibility for students who do some research work on these materials to actually publish their report or something like that, which would actually make it possible for people to access to see what other people have found out?
Erway: Absolutely. We found with The AMICO Library™, where we have had more years of experience, that whether people actually used the model assignments—which showed where there was content in the collection to support a particular unit of art appreciation, for instance—or not, it at least gave them an idea of what could be done with the material.
You are talking about something more sophisticated, basically adding to the database publications that were somehow informed from the materials in the database, which is certainly of interest.
Copyright Restrictions
Participant 7: In my library we have tried to digitize a great part of our poster collection mainly for preservation. As an RLG member we decided first to put that data in the RLG Cultural Materials database, and suddenly we realized that we were not allowed to do that because of copyright reasons.
Do you intend to have all copyright-free material in that database ordo you intend to tackle the problem of copyright with that project, whichcould be quite difficult to solve but quite interesting for all of us tosee what solution we could have to this problem?
Erway: To date we have tackled it to a certain extent through the agreements that the alliance participants have signed. This is the easy part of the rights issue, where contributors say, "We will only give you things that are either in the public domain or for which we control the rights or for which we have received permission to do this."
I would not say that only things in the public domain would be in thedatabase, but hopefully only things that are legal to have in the databasewould be there.
On the other hand, we have some alliance contributors who are using this as an opportunity to provide access to collections they don't feel they can put on their own open Web site. Fair use in the United States is not universal, but there is a comfort level at having a database for which the use is restricted to research and educational use. Some people are saying, "I can't put that on my own Web site, but I feel much better about putting it in this secure database."
The other possibility is to put things in with the proviso that, if someone notifies us that there is an ownership issue, it will be removed. There are very seldom lawsuits if you take the approach, "We have done the best we can to make sure this is okay. If it is not okay, tell us and we will make it better."
A Place for Popular Culture?
Participant 8: On the American Memory site there is a terrific collection of baseball cards which I am sure attracts a lot of kids but might not contribute very much to classroom assignments. Playing cards, stamps, and so forth may be tremendously interesting, but might not support the purpose of a researcher at a university.
Participant 1: They are certainly of interest at Harvard, where we teach baseball!
Erway: Which is exactly why we allow you to do the selection. We put in whatever you give us. You know what is wanted.
Participant 9: Earlier someone mentioned political buttons as an area of political literature that people associate with elections. I don't know if the buttons themselves would be copyrighted, but there is so much pictorial and social history there that would go with some of the other things that have been mentioned.
Moderator: Or visit any presidential library and see what the American citizens have donated to their presidents and have that all on the Web.
Preliminary Analysis of Use
Participant 10: Can you tell us about the use analysis to date?
Erway: We have data from January through March. Afghanistan is one of the top five searches for which we return no results. Marilyn Monroe is toward the top of that list, too.
The collections that are most heavily used are, not surprisingly, the collections that are best described. There were a lot of things that would have been relevant to the searches in my demonstration that I could not retrieve because they didn't have very many words in their descriptive records. The Duke sheet music records are rich. There is a lot of description and a lot of subject headings.
What will be used is what is returned in the result set, which is largely dependent on how rich the records are.
Participant 11: That is very helpful. Some of the analysis can be regularly fed back to people who are supplying the data so that they can make informed decisions about what they are doing.
Moderator: Especially the browsing or surfing behaviors. I would love to do some studies or see some studies done on the knowledge that people are taking from the information we provide.
Erway: I think our first priority is figuring out how to let people know that this exists and all the different ways it might be used, and encouraging faculty. It is one thing to subscribe; it is another thing to get people using it. The first thing is how we get it into the hands of the people who want it, and one of the ways to do that is to show what others have done with it.
Participant 12: This is an opportunity to plug the user study that we are doing at Penn State, which is a study of our population of 2,000 academics and their use of pictures in the arts and humanities and social sciences. People continually use transaction logs from our legal library to try to help determine what content should be there. Is there any interest in doing so on the part of the contributing institutions?
Erway: I am not sure how much time the AMICO museums spend analyzing usage data. They get reports of who is using their images and which images, as well as search results.
AMICO has implemented a way that an end user could say, "I looked for this and didn't find it. I wish there was more of this in the database," and the AMICO members do respond to that. Any time there is a request for an image from a museum that is known to own the original, they are really responsive to getting that content in the database.
The AMICO Library is a case where RLG doesn't have a direct connection to the contributing museums. We will have more ability to provide that kind of feedback to the contributors to RLG Cultural Materials.
Broadening Use
Participant 13: How can you motivate those who can't afford to subscribe to contribute their content?
Erway: Right now we are looking at the next audience for RLG Cultural Materials, which would take it beyond institutional subscriptions and make it more broadly available to individuals. We imagine there might be a freely available, smaller version of the resource, maybe thumbnails and brief captions. When you find something you want to use, you tell the system how you want to use it. Are you going to publish it on the cover of a journal or are you going to put it on your Web site or do you just want to print it out and put it on your wall? There will be a pay-per-view opportunity, and that money would go back to the contributors.
Part of the problem for us is that the service needs to be self-sustaining. As with RLG Archival Resources, we have a lot of contributors and a lot of subscribers, and I think all of the subscribers are also contributors. If you offer a really great deal to motivate contributors to subscribe then you have to charge some horrific amount for people who don't contribute.
Participant 14: I'm wondering about the earlier observations about allowing people to make comments, allowing people to make recommendations about what is related, allowing virtual communities to develop, and allowing personalization capabilities. Some people are a bit hesitant about it even when it is their own anointed communities. Do you start to feel differently about it when you have a much wider audience?
Moderator: Isn't that what our mission is? The wider audience we can get, the better?
Printable Version
