 Welcome. I'm Rick Ferrante. I'm with the Smithsonian Station Archives. I'm in charge of the digital services division, which includes digitization, revenue media outreach, and the electronic records program, which is focused on born digital material. My co-presenter, Rusty Russell. I'm the program director for collections and informatics, collections being actual dead plants in our reference collection at the Department of Botany at the Natural History Museum. And before we get underway, we want to talk or we want to ask a couple of different things. One is please tweet away, but if you have your volume turned up, if you could put your phones on vibrate, they're recording, we don't want fast and come up. And then we'd like to get a sense of what roles you have at your organizations. So if you kind of go around the room and introduce yourself and say what it is that you do. Let's do it a little bit more quickly. Let me try this. It was one of my library, the metadata made in the Smithsonian, who was interested in what the makeup of this audience would be, because the subject matter maybe is just a little bit non-mainstream. How many would qualify themselves as an IT person? And how many would qualify yourself as a library slash archives type? And then any faculty type roles? So we're here to talk about a project that we've been involved in, which is having the effect of extending the life cycle of quite old material, which was not only analog but handwritten or drawn, and so it's been hard to access. So we like to think of this as buried treasure. So these materials, a lot of which we have in our collections, date back to the early 19th century, and are therefore hard to get at, even more so if you don't know what you have. If they haven't been catalogued, if you don't know quite which corner they're in, makes it relatively impossible, unless you know somebody who knows somebody who knows somebody, to include that material in your research. So that is part of what we wanted to address. We wanted to create access to get past that issue. If we can do it in a way that makes it machine accessible, then all of a sudden it becomes incredibly useful for people who are working with today's tools. Until that happens, the information life cycle is pretty limited. So this is a wonderful graphic to simplify what it is that we have buried treasure somewhere underneath the sidewalk. So what I just described is that the amount of access that something has, particularly analog, a couple of centuries old, limits the value that it has. So in 2008, Rusty and Ann Van Kamp, the director of the institution archives, kicked off a project to identify this material, to locate it and to provide access through an online catalog with initial funding from Clear. So they broke ground, so to speak, and we've made a lot of headway. Since then we've identified, located and identified and cataloged 9,000 fieldbooks to date. We're using mods, EAC, CPF, and natural collections description, another standard to make this information, not only accessible, but interoperable. Along with that, we've begun to digitize, made much faster progress cataloging than we have digitization. On average, this material is 100 to 200 pages in length. So it's not like photographs, it takes a little bit more time. But we want to go somewhere else from there, we don't want that to be the end point. And so we teamed up with the Smithsonian Institution Libraries in 2013, kind of expanded the group. Because of how we made these things initially available, we've been able to expand access through aggregators like Digital Public Library of America and most recently the Biodiversity Heritage Library, which is really good. And that's kind of a new set of collections for them in that it's primary source material as opposed to published material. So many hands make for light work and new depths. So we have online catalog, we have digital material available, we have access through aggregators. We wanted to get inside the material. That's what people were asking us for. They didn't want to just see it online. They were glad that they were, they now knew that it existed, but how do they bring their tools to bear? Transcription solves that problem. We don't have the staff to transcribe 9,000 fieldbooks and we're still cataloging, so that will continue to grow. So we got in when the Smithsonian was starting to build a transcription center, a crowdsource transcription center. And we've been able to take advantage of that. The crowd is coming and looking at these fieldbooks, developing relationships with this material, and has done an amazing level of work for us. So what this allows us to do once we identify the quality of the material that we're getting from them, the transcriptions, is to get full text indexing off of this material. Allow it to not only be searched at the catalog record level, but material insight of that, that can be made available as XML based data sets. Then you can bring tools to bear with that. Rusty will talk a little bit more about how these are used. This is an example of a record being searched. And then I searched for Mill Canyon, and it's one of multiple fieldbooks, but it's showing the pages where that, actually, that locale is identified. And you can click on one of those and it jumps and takes you directly to that transcription. So it is a great way to not only get inside one particular object, but to identify multiple objects that are dealing with that locale. And as you add additional terms, you get more granularity. So as you can imagine, it creates all kinds of opportunities. Just because we can do something doesn't mean it's worth doing. And the concept behind the fieldbook project wasn't simply an exercise in finding fieldbooks in the institution, which it was, because we had no idea what we had, for the most part. And yet, these are the original sources of information that are attached to the collections that we maintain. It wasn't either about making them digitally available. It was about answering the questions, so what? And I'll get to that in a little bit. As a field biologist, I'm feeling a little bit like a fish out of water here. One of the things that we're used to is doing the fieldwork that we do. It seems a little bit romantic at times. But the words that I think are most powerful for so many people is explore and discover. And this is the beginning of the acquisition of data. In the past, expeditions sailed around the world looking for natural history collections. You know about the U.S. Exploring Expedition, the Challenger, Captain Cook, and so forth. Also, Lewis and Clark, a fairly famous story for North Americans. And Charles Darwin, Alfred Russell Wallace, and Livingston went out on these romantic explorations to discover. But as a field biologist, you spend the time afterwards investigating the objects, observing, surveying. I had nice backgrounds here, by the way, but he kind of stole them. And ultimately, to document what you have, and then even more ultimately, to record for posterity. This ends up, for the most part, in the field books that, as field biologists went into the field, they recorded it. The last 20 years or so, we've seen a move towards digital acquisition of data in the field. But the vast majority of information like this is locked up in analog objects. I started a project to find all of the relevant material from the U.S. Exploring Expedition, which was a government-sponsored expedition four year round the world voyage, collecting natural history objects along the way. And ended up visiting more than 16 institutions, libraries, archives, historical societies, and so forth. And still haven't found everything. And it occurred to me that, now, you can build a silo of information. But there are other silos of information. There's a lot of connected material that are relevant to one another. And at the moment, all we were doing was building silos. So another longer-range plan for the field book project was to begin to communicate between these silos of information about field books. And to be able to answer questions like, well, do you have so-and-so's field books? Well, you might have three, they might have two, and so forth. This is one of the journals from the U.S. Exploring Expedition. And here's where the rubber hits the road. The specimen that is referred to on this page is one that has information that includes only the name, the locality, Oregon, and the number 848. Oregon's a big place. That's a big dot on the map, knowing more precisely where that is. And knowing when that specimen was collected helps us create information that, I'll get to in a minute, I call the three dimensions of biodiversity. So there's a lot of this kind of content. And we're finding it, we're digitally imaging it, we're transcribing it, we're making all this available, but so what? I started using this concept called the three dimensions of biodiversity when I introduce visitors to the herbarium. Our herbarium, for those of you that don't know, is a collection of dead plants. It's a reference collection. Each one is a record of that thing being collected at that time in that place. It's a point in the three dimensions of biodiversity. So that thing, that species, in another place starts to create dimension. And the five million specimens that we have with all the localities and all the dates create this incredibly rich model of biodiversity worldwide and over time. But the 120 million of them just in the United States would create an even greater picture. And the hundreds of millions around the world, you can see where we're going with this. The problem is that none of it is digitized. So I talk about the organism being in a particular place at a particular time. Unfortunately, that kind of information isn't always present with the specimens, especially with the older ones. So having access to the original field books in which most of the data was never extracted to join the specimen or even to find its way into publication is really critical. Let's look at it in a different way. This is a field book, again, from the U.S. Exploring Expedition, one of the theologists. And in there, if that were a botanical field book and it's not, there would be a reference to a plant specimen that was collected. Ultimately, information like that finds its way to the literature. Now, as a scientist, I'm interested in all of these things. But my entry point could be any of the three. I might have a specimen in hand and I'd like to know where the field book is that reports the acquisition of that specimen. Or I might be looking up a publication and want to go back to the actual specimen that is referred to there. I need to know where it is and all of that. So back to the beginning when I was using different words to describe what we do, field books are about exploring and documenting. Specimens are the acquisition of collections in the field and preserving them for posterity. And finally, publication is the report, the study. So there are different ways to look at this but you see that this issue of what we call connecting content is really critical. So that regardless of your entry point, you have access to all of these things which are inherently related. So a big part of the field book project then, the next level of the vision had to do with not simply having our silos talk to one another about field books. It was about connecting to other kinds of content like the published literature through DPLA, Biodiversity Heritage Library and others. And also to collections at natural history museums, universities, other institutes all over the world. Most of which have not been digitally imaged. Most of which have not had the data collected from that particular specimen. But as we start to create that, again we're creating silos of collections, silos of publication data. But here's where it gets really dimensionally interesting. And that is in this field book, which could be a simple listing or it could be a journal type of field book, there's other kinds of information that don't fall into this really simple model of biodiversity. There's other information that would be very likely interesting to a social scientist or to a meteorologist or to other forms of study because as individuals were going, they were recombining on their surroundings. They were referring to information that they were observing that had nothing to do with why they were there and yet they were still recording this. And that kind of information being extracted from these field books through word searchability became that much more important and the biggest part of the vision. So the problem was we could get a digital image. Could we get it to the point where we could transcribe it, mark it up and make it available so that people in all disciplines, people in all endeavors of study would be able to have access to that. There's information in these field books that I've read regarding the behavior, the behavior of animals they were collecting. The symbiotic relationship between various organisms that they observed in the field, the names, and this is primarily the names of the organisms which we have a habit of changing over time as our concepts of species limits change, information about ethnobiology, how the plants and animals were being used by local cultures, the images that are present in field books through sketches and illustrations and that sort of thing. The traits, the characters of collections that are made, the DNA sequences that ultimately result from the testing of individual specimens, that information relates back to the individual specimen and gets connected to everything else where those connections are possible. So, within a particular field book like this one, we have the what, where, and when. We have the simple three dimensions of biodiversity. But again, we'll have all of these different data present. Junius Henderson was a paleobiologist at the University of Colorado whose field books were transcribed a number of years ago and the marking up of that and the making all that information available was actually funded by NOAA because of the weather implications he was very good at recording weather data. So there's this plight that most scientists complain about and that is there's never enough data. You could have all you can have, there's just never enough data. You always crave that much more. So the kind of information we have here and generally the results, the analyses that you produce in most studies are based on a subset of available material and it may not probably isn't a complete subset or complete set. So this concept that we'll never have enough is a problem but one of the things we're attempting to do with the field book project is to satisfy that in the areas where we know there's information, we know how broadly it is, how broadly useful it is. The Taxonomic Databases Working Group is a very old organization that began as a botanical study group back in 1986 and since then has become an organization called Biodiversity Information Standards and still meets on an annual basis and has a really brilliant combination of traditional field biologists and high tech information savvy individuals. And two of the things which at the last meeting in Jantroping, Sweden, I just learned how to pronounce that, that we're really going to begin focusing on is to accelerate data acquisition. We can't go fast enough. There is so much that if we plot along at the rate that we are, we're getting behind because we're creating data faster than we're making it available and to facilitate content delivery. So again, through the field book project, one of our goals was to do exactly that. So at this point in time, we wanted to allow for plenty of time to have the next step, the conversation. One, how does our experience inform your ambitions? How does your experience help us refine arms? And what are the possibilities that we will identify through the course of the conversation? Well, no, I take that back. You have to have the capacity to read handwriting. And trust me, that's a skill set that slowly, maybe too rapidly, we're going to begin losing. There are a lot of nuances in these field books that are difficult to get your head around, a lot of abbreviation of things that once you know it, it makes sense, but not knowing it initially creates confusion. But a lot of our transcribers, most of whom are anonymous from all over the world because we can track their machines, are doing it from... We don't know if they're professional. I love going on there being OCD, going in there and you start transcribing things and you can't stop. But really, we are putting it out there for the community at large. And those that we do know, those that we have identified, tend to be those people who are really interested in the subject, but have never had an opportunity to contribute, to give back to what they think and as a result, they tend to be in the later years of life too. So we've done a little bit more. I've had a couple of fellows work on the community management side of things. The way the transcription center is designed is you can transcribe without identifying yourself. You can do part of a page. You can do multiple pages. You can mark it as ready for review. If you register, if you tell us your email address and create a name for yourself, then we give you the privileges to review. So you can review what other people have done and tell us whether or not it's ready for the institution to review. So there's that three-level process. There's no algorithmic back-end to it. We're working through that process. And what we're finding is that we are getting incredibly high quality that people who are doing the review really care about this. We think that there's some brand affiliation and the ones that we've been able to have conversations with have told us we like contributing to the Smithsonian and knowing that we are helping the Smithsonian do its job. They have also said that they like bouncing around. They will come and they'll start to work on a particular project and kind of get full of that experience but not want to stop. So they bounce to another project. So we have botany specimens. We have field books. We have bumblebees. So they're helping us create new digital catalog records for some of this material. And they'll bounce back and forth. And they really enjoy that. We originally thought that it would be professionals with special areas of expertise that would be the only ones to really engage. We found that there is another constituency, if you will, who came for the novelty, discovered a passion, stuck with it and then through their social media presence have encouraged their network to come and participate. So we've really seen this develop in a lot of different ways. Another one of our transcription strategies for the botanical collections that are up there because I also have a foot in that place as well is coordinating with professors in different places to do content that is germane to that individual's research and then they have their students do the transcription and that adds a lot of depth to the, it provides the data to the individual researcher but it creates an additional bonding between the students and the professor. The completed projects are available in a number of different ways. There's that learning experience. One of the other questions. We have, Cleo Calderon was a botanist who worked closely with Tom Soderstrom, particularly on bamboo-related species in Brazil and northern South America. And she wrote for herself as people do. And she was multilingual so depending we think on who she was talking to last, her notes were in Portuguese or English, sometimes on the same page. And that took a little bit longer but we were able to call out and say, hey if you speak and read Portuguese we'd love your help with this particular project. What's interesting about that, Cleo, I collected in Brazil with both Cleo and Tom and she was Argentinian so her first language was Spanish. She was extremely good at English but she was especially good at Portuguese and all three of those languages show up in her field books very randomly sometimes. Also an interesting story that she recollected a species that had not been seen in the wild for more than 100 years. We were actually out there looking for that and it was her team, not ours, that ended up finding that and her record of this, it was called Anomacloa, her record of this was very understating. I went back to look for it because I thought well this would be really interesting. I want to see what Cleo said about it while she was there looking at this thing. Found it or whatever, I forget exactly what she said but it was very matter of fact. And that was kind of interesting to me because she was otherwise a crazy hot-blooded Argentinian. I had to say a question about the translation. So this started in 2013, the collaboration between the Smithsonian and the field of policy projects? The Natural History Museum, Department of Botany and the Institution of Archives embarked on this in 2008. That's when we got the initial clear funding. It's continued since then. In 2013, the institution libraries came on board. They're a separate group as well so additional resources, additional avenues and we've been able to really kind of explode the ways into this material since then. It also provided stability because Ann and I were constantly looking for funding to keep it going and having it under the umbrella of the libraries and BHL made it more sustainable. We're over 9,000 now and because we've found all the low-hanging fruit, you have to understand that natural history curators are the archivists of the person who preceded them. So they're hanging on to this because they're responsible for protecting it and their descendant will do the same for theirs. So there's a lot of content still in people's offices. There's a lot of objects in boxes somewhere. So as we find them, as we locate them, we're working on it. My guess is that we've probably got another 3,000 to 4,000 to locate within the institution. Yeah, we have an ambition at this point in time to, over the next couple of years, digitize another 2,600 field books while simultaneously cataloging 2,000 to 3,000 as quickly as they can be digitized. The appetite of the crowd to transcribe more than satisfies that. So the transcription center has been around for a year. We are halfway through our backlog of digitized material in terms of the transcription center. We do have to exercise some judgment. Some of the material doesn't have enough contrast. The pencils were too light in the field or the ink is faded too much to actually really be a successful experience for somebody from the crowd. Grad students or scientists are going to be much more interested in making headway with that than somebody who's there for fun. But the large bulk of what we have is appropriate for the transcription center. While it won't keep pace, we've gotten through 16,000 images today in the year and six months that the transcription center has been available. The field book documents the acquisition of these, whether we have them or not is a different issue. We may have 20 plus contributors for a single page. Until it's marked ready for review, anybody who comes to that page, there's an editing lock on it so while somebody's working on it and somebody else can't, but they can go back and forth. The whole idea of simply saving it and not marking it for review is, I'm not confident or I didn't finish the whole page, I have questions. There's instructions we give them to identify where they have questions and an area where they can write notes to each other. Back and forth and we do see people who've found great sources for specimen names or Chinese translations, we've seen them do wonderful things there. When things are marked ready for review, the registered users are able to go in. They really care about accuracy and sometimes they will go back and before we have a chance as staff to go in and review their review, one of the other reviewers will say, no, this was marked ready for approval before it was actually complete and they will ask us to reopen it for editing. We have had different units approach the Smithsonian level approval different ways. In some cases where they're working inside a journal or a diary or a set of letters, one particular unit had fellows who as part of their research did that Smithsonian level approval. In other cases, we have particularly the people who are documenting the specimen records and they need a higher level of accuracy. Something that's 99% is not good enough, particularly because of where this information is going to get used. It's a lot more intensive on their part. How they source that is much more difficult and it's definitely a resource issue because you're creating records so much faster than you normally have staff allocated for which is a juggling effort that Rusty could probably speak to better than all. Because I don't actually engage in that, all I can tell you is that those individuals in the department that we have reviewing content are fully engaged in that and we're struggling keeping our head above water. But Jennifer, you touched on one of two basic strategies for doing that. If you have ever heard of Zooniverse, it's a crowdsourcing facilitator and their strategy is to have multiple people do the very same thing and then match them using various comparisons or algorithms to see what the final product is going to be but then it still needs to be reviewed. We're doing it a slightly different way by having as many people as want to contribute to this particular record do that and basically do what people are very good at doing and that is catch you at making a mistake. So somebody didn't get this right, somebody else corrects it and you just go through these iterations until you get to the point where somebody finally marks it, then our reviewers are the first level of review that takes the onus off us at least at that level. So you have two options like that and there are others but both of them have a different approach to addressing the ultimate goal and that is get it right, have it completely accurate the first time. Then this is harder to do often times with plants or with actual specimens because you've got, although it's true also in field books, a lot of scientific terminology, words that you're not familiar with, Latin and Greek, for example, in the nomenclature and abbreviations of that, PAPH, Pathiopidillum, the orchid, which is PAPH, nobody would understand that until after a while you've actually seen that in that particular field. So there's a lot of that we have to struggle through. The idea that out of sometimes 20 people working on a single record, one person is going to know what that abbreviation means and so finally it gets to the reviewer in pretty darn good shape. Some of the other examples of issues are geopolitical, right? Think of China in the 19th century. Those place names don't exist in many cases, completely different. Country boundaries have changed and so what's actually in the old material would find some of that resources in order to address those issues within the material. Multiple units approach the transcription center differently. We do. We try and identify when something is particularly easy or particularly difficult. I have mischaracterized a project and quickly heard back that it was more difficult than I had led them to believe. We've also found that the people who develop the passion for transcription will talk to each other, particularly in Twitter. They call themselves volume peers and they will tell each other whether or not something is particularly easy, whether or not something is a good project for people who want to develop review skills to work on. They will actually ask people who have foreign language specialties to engage in a particular project. They're calling each other and calling for specific skills to a project even when we haven't quite done that on the front end. It's a great relationship but yes, we found that that's very helpful and the easy ones go really quickly. Another thing that we do to attract interest in some of these where since we do have five million plants to do, just doing them all and throwing them up there in some random fashion doesn't attract the attention necessarily. So we will put things up in packages. Like we did with the Malvase, we did it as hibiscus. The Tiliaceae, we did it as basswood. We're putting together little packages to attract people in that fashion. One of the things that we're talking about is doing massive digitization and then filtering out various ones. For example, all those with a Cyrillic language on the label get ferreted out to that kind of content. A lot of the printed labels are going through OCR right now and then that gets thrown up as just needing to be edited. Initially we hired a project manager, we hired a cataloger and we hired a cataloging coordinator and we hired a cataloger. The first grant was half a million dollars to work for about three years pulling all that together. Our original goal of also creating an online application that would achieve our vision for world domination which was an international field book registry didn't quite make it to that point. We had subsequent grants to do digitization, to do conservation which by the way I think we need to point out we're talking about fairly fragile old objects here and of course those people who are librarians and archivists appreciate the unique value of these things as opposed to being one of many but also the fragile nature of them and many of them need to go through some form of conservation or object remediation before they can even be held and cataloged. So that wasn't money we had in the original half million. So we were getting additional grants like that. We struggled through another couple of years and then our conversations with the libraries and BHL brought us to the point of having a more sustainable model for continuing this thing forward. We're in the process we're waiting on, forever waiting on the next grant that has been promised to us and that is paying our new project manager, Julia Blaise, sitting back here. And the staffing is pretty much the same. We have a cataloging capacity through the archives. We have conservation and digitization imaging, digital imaging being done. So those are basically the various groupings of functions that need to be performed. So with regards to conservation and digitization we found as we were going through the cataloging process that if we were going to proceed any further roughly one out of four field books needed conservation. Some very significant interventions and others and not so much before they were stable enough to be digitized. And so the additional grants that we were able to provide one from Save America's Treasures allowed us to bring in pre-doctoral fellows and some cases some post docs to work under the auspices of our paper conservators on that material. I allocated a portion of my digitization staff to do digitization because as soon as people were aware of it they wanted to see it but they couldn't get to DC and supplemented that with interns and volunteers. QA process was always paid staff with regards to the transcription center. We rolled into our overall workflow. So we're digitizing as a matter of preservation and it's part of our ordinary operations. We take the output of that. Our endpoint for that is structured in a way so that we have just a little bit more workflow to do in order to put it into the transcription center. Roughly we're able to keep pace with that plus doing metrics with roughly six to eight hours a week of effort in that. Approval would be another couple of days worth of effort if we were going to be doing that. That goes a lot slower because we don't have the additional staff for that. As I've heard in lots of other presentations we as a community have the obligation to advocate the value of what we're doing and why it is appropriate to invest additional resources but we're definitely seeing that happen here and as more and more researchers are accessing and using the material they're becoming advocates as well. One of the things I want to mention too is the use of social media in integrating social media into the fieldbook project. In the very early months, Carolyn, Senoi, Leslie had come on board. Senoi came to me and said, we would like you to contribute a blog, blog post for the blog and I said, we have a blog. They had gone ahead being the 20-somethings that they are and created the most amazing blog on fieldbooks and our content and the stories that they're telling. It didn't have as much to do with science as it had to do with people and stories and personal issues and they were so good at bringing stories to light on anniversaries of various sorts. Edgar Mearns took his son into the field and we posted that on Father's Day. All of that, there was a poem, a love poem, that somebody wrote that went up on Valentine's Day but you read this and it's the most, it's the one thing that more people will comment on that, oh, I really enjoyed your blog. Oh, by the way, you're doing 9,000 fieldbooks. That's great, but I really liked your blog. The stories are compelling. Some of them are life and death and there's lots of imagery and that sort of thing. So that gave us a fair amount of visibility in ways that was one story about that. We got a contact from the son of a Kenai elder in Kodiak Island in Alaska who noticed the story that we had put up about William Fisher who retired in the late 19th century and found a Russian wife on Kodiak and lived there but he was really interested in ethnobiology, did a lot of collecting. We have some of his plant and animal specimens and he had just seen this story on William Fisher and he also saw that we had at that point digitally imaged a number of his collections and he said he shared this with his mother because by the way, Fisher also collected Russian names for that plant, Kenai names and well, English names. Common names. Well, common names too. But she was willing to look at these things and give us the appropriate updated pronunciation and spelling for these plant specimens that we had and they were really excited about that to be able to contribute in that way. The blog was a very serendipitous thing that was out there that he saw and created a really interesting, so we blogged about it, but a really interesting contact. Some of the things that surfaced out of this are relationships between scientists that aren't necessarily captured in the individual objects but as they tell their story about this, Frederick Coville doing his work in Death Valley in 1891 turns out that he went on a number of expeditions with a guy named Vernon Bailey and Seahart Marion and Seahart's sister Florence was an ornithologist as well and seemed to tag along on some of these and Florence and Vernon hit it off and later became an item. And in the course of this we are finding some of the people who have come to the transcription center go in and look for secondary contributors that aren't necessarily in the published media or in the case of early or 19th century are referred to by their husbands' names. Mrs. Sidney Blake actually turns out to be Doris Holmes Blake a collector, but how the convention worked then doesn't necessarily bring that to the surface. So we're seeing a whole another set of authorities come out of this material. One of the things I clamored for at the very, very beginning was we need more information, more keywords about each object because, sorry, but an archival finding guide is not helping me. There are field books in those RU units that nobody knows about because it's just not described quite that way. So I wanted information about the people, information about the localities, information about the content. And so that created the data model that we initially created incorporated a lot of that. And part of the cataloging practice we're using multiple thesauri with regards to that but to stress the keywords because of the search engines that we now know people go to those keywords and embedding them as meta tags is really important because we have quite a number of field books that were field notes 1891. Because they weren't thinking about what somebody else might want to look at that. And the only way to narrow that set of material for what you're interested in is in that set of keywords. Google is going to look to that information not necessarily dig into what is in your subject index terms and so on. Right, exactly. Finding A is actually as Rusty was describing is less useful than an item level catalog record because you only have that object or that folder description and that doesn't give you the rich access points that you would get from keywords. We found a watercolor book in a unit that was described simply as OF Cook in Liberia that was done by one of his protégés, Frederick Straub, who died as a 27-year-old collecting in Liberia. Beautiful watercolors of plants and next to each one was a little number which it turns out linked back to plant specimens that we had in the herbarium. He collected with that number. Of course, if you know anything about pressed plant specimens they're not very colorful sometimes. And to have these illustrations adjacent to the actual specimen was, and finding it though, we found it. You bump into those sorts of things. They're not clear in the original finding. If I could kind of shift a little bit and talk about the community that's come to help us do this. We give back to them. We've done a number of Google Hangouts giving them a virtual behind the scenes. Talking to people like Rusty and other curators about an aspect of the material that may be available in the transcription center. And it's really, really well-received. We encourage them to help us complete a project in order to get access to a contributing connect session. We have seven-day review challenges and then we've taken to recently a monthly, we call it MyTC Discovery, where they share with each other through social media the discoveries that they make. Which helps us make connections that we wouldn't ordinarily make. So everybody benefits from that. But that's, again, a way of... We're not... It's easy to think of crowdsourcing transcription as a way to get free labor. We want to think of it more as a way to, as a partnership, create new knowledge. And they really receive that well as being part of the group that's creating this new knowledge. It means that we have to demonstrate a respect for what they're contributing and the intelligence that they bring to the process. We try hard to do that. There is an effort associated with that. If you Google the fieldbook project, that'll be your first hit. You can read all about it. And you'll see the other e-mails or websites up here. If you just Google Rusty and the Smithsonian, you'll find mine, too, somewhere. It's pretty much... And I'm happy to hear from anybody with additional questions if you think about them later. Thanks for coming. Thank you.