My project at the Missouri Botanical Garden focuses on access to illustrations in BHL’s corpus of biodiversity literature. I’ve dipped my toes into the related areas of interface design for digital special collections exhibitions, digital humanities, metadata, social media outreach, and rare books in the course of my studies and work. The possibility for engagement and exploration of cultural heritage in the digital environment is infinitely exhilarating. I am fortunate to be able to dive into these topics while making concrete progress on a project that will serve others. I am also lucky to benefit from the knowledge and experience of my project mentors, Trish Rose-Sandler (Data Analyst, Center for Biodiversity Informatics) and Doug Holland (Director, Peter H. Raven Library.)
The deliverables for my delightfully named “Treasures Unlocked” project build upon the primary intention of Art of Life grant project led by the Missouri Botanical Garden: A way to discover illustrations through the BHL portal. I am responsible for technical requirements and interface prototypes, based upon user studies, as well as a report on image discovery best practices. In 2012, there were more than 47 million digitized pages in BHL. With the support of an NEH Humanities Collections and Reference Resources grant, Trish Rose-Sandler at the Missouri Botanical Garden served as principal investigator for the Art of Life project. This was primarily an effort to identify and describe illustrations among these pages for eventual discovery in the BHL portal. Metadata was crowdsourced mainly through Flickr and Science Gossip, according to a new schema created for the project (Rose-Sandler, Crowdcon 2015 Proceedings, p.39.) A secondary goal grew in importance as the project progressed: to continue to expose new audiences to BHL illustrations through these exterior platforms. Additionally, the project led to the formation of a particularly successful research community, Science Gossip (Rose-Sandler, Crowdcon 2015 Proceedings, p.37).
These last few weeks, I’ve taken advantage of the resources at the Missouri Botanical Garden and BHL to gain contextual information for my project. The Missouri Botanical Garden’s staff in Research and Conservation, books and exhibition catalogues about historic natural history illustrations, and BHL’s Outreach and Communication manager, Grace Costantino, and her work have been very helpful in this regard! I have gained an introduction to:
- Systematics and other fields related to the study of biodiversity
- Natural history illustration techniques, from historic to the current day, and their role in the taxonomic publishing process
- History of digital libraries at the Missouri Botanical Garden
- Discovery and usage of illustrations among systematic botanists
- Designated BHL user groups and related outreach efforts
- Intentions and outcomes of the Art of Life project
At this point, I am looking more deeply at conversations regarding citizen science and crowdsourcing of cultural heritage, and considering how the various models fit into the mission of institutions or digital libraries like BHL. I am also beginning to look into the realm of user experience to plan my approach for better understanding, reaching, and gaining feedback from those who would use such a discovery mechanism in the BHL portal. In some senses, my role is that of a data curator, since I hope to find appropriate ways to help manage and share the digitized illustrations and metadata produced by crowdsourcing throughout their usefulness to scholarship. For more on this approach and a similar project to my own, check out DHCuration.org and this blog post by Lydia Zvyagintseva about developing prototypes using crowdsourced metadata from NYPL’s What’s on the Menu project.
One of the primary tensions that I have encountered in my project is audience: BHL considers its core users to be scientists and librarians. Yet scientists may have limited use for historic illustrations created through traditional printing techniques, the main types of illustrations for which we have crowdsourced metadata. According to Jim Miller, Vice President of Research and Conservation at the Missouri Botanical Garden and systematic botanist, historic illustrations are referred to in relatively rare cases, when:
- An illustration has been chosen as lectotype (when a type specimen is unavailable.) This could be a substantial number of cases, since the number of plant species is estimated at up to 450,000. This substitution happens perhaps once in several thousand times.
- An illustration can provide clarification when the original written description, or protologue lacks sufficient detail.
Illustrations have, and continue to serve as an important method of communication among scientists in biodiversity related fields. In the case of taxonomic literature, it is ideal to include the work of a scientific illustrator highlighting the structural features of a species in high detail. According to the botanists I have spoken to, even high quality digital scans of dried specimens and digital photographs of live plants do not achieve the clarity of illustrations, but are a useful alternative. Maps, charts, and photographs are also often included in publications relevant to biodiversity. These can be created through a variety of means, from pen and paper to the use of digital technologies. In their recent article, Willi Eglof et al. provide more detail about the various forms of illustrations in taxonomic literature. Based upon manual classification work on a subset of images for Art of Life, the most accurate estimate I can provide regarding the forms of images in the BHL corpus are 52% drawings, paintings, or diagrams, 24% photos, 19% charts or tables, 4% maps, and <1% bookplates.
BHL’s efforts to raise awareness of its offerings and build engagement surrounding them among artists, bibliophiles, and citizen scientists have demonstrated these groups’ interest in historic illustrations. User engagement is both a strategic goal and a commitment to BHL’s values to “facilitate dialogue about scientific issues within social and cultural contexts” and “stimulate curiosity about nature and culture, inspiring people to learn, explore, and better understand their role and impact in this world.” Flickr tagging for Art of Life (AoL) built upon BHL’s existing Flickr presence and encouragement of taxonomic tagging. The BHL Flickr site had been established about a year prior to Art of Life. A few months before AoL began, BHL celebrated the engagement it fostered. Grace Constantino explained in a blog post that “Reaching new audiences and demonstrating that BHL has much to offer outside the realm of taxonomy is a critical step in the growth of the project.” While BHL’s social media presence on Flickr, Facebook, Twitter, and the BHL blog are managed by Grace, citizen scientist Michelle Marshall also shares BHL illustrations through Pinterest (started July 2012), Instagram (August 2016,) and Tumblr (October 2016) on a volunteer basis.
The Science Gossip (SG) community was described by the AoL team as “the biggest and most unexpected success” of the Art of Life project (Rose-Sandler, Final Project Report, July 2015.) SG was born of a competitive offer for development funding and research expertise from the Constructing Scientific Communities (ConSciCom) grant project. The Missouri Botanical Garden team members worked with ConSciCom and Zooniverse to develop the custom project site. In my recent conversation with Dr. Geoff Belknap, the lead academic researcher involved in Science Gossip, he explained its focus on developing a research culture surrounding Victorian periodicals. In addition to creating image metadata, the group pursues and shares interesting discoveries about the materials. He described this model as allowing people to learn through the archive. His role as an academic is to frame or reinforce valuable questions.
Understanding the relationship between BHL and the people involved in these two different crowdsourcing models will be a major focus of my work. I am looking forward to getting to understand the motivations of the Flickr and Science Gossip groups, and their viewpoints surrounding the use of their work for portal enhancements. I would like to better understand the motivations of those who interact with illustrations through BHL’s outreach efforts. I would also like to understand the needs of communications, exhibition, and education staff especially at BHL institutions.
Finally, another issue that I am considering is the scalability of metadata creation, and how it fits into the scope of my work. February’s ingest from the BHL Flickr API and up to the minute stats from Science Gossip, compared to a recent query of the BHL database reveal that at maximum, less than 1% of the 4 million illustrations in BHL have any sort of metadata beyond illustration type assigned to them. Through BHL’s Flickr site, 31,771 illustrations have been tagged with with 134,176 annotations. Through Science Gossip, a maximum of 152,604 pages have been annotated 544,504 times (up to 4/5 may not contain illustrations.) Flickr tagging activity increases when calls are placed out regarding campaigns about special materials, like the Joseph Dalton Hooker exhibition at Kew Gardens, but the current rate of Flickr and Science Gossip metadata creation will not allow us to fully describe these materials anytime soon! Analyzing and prioritizing the remaining images and looking into computer-assisted description are certainly tempting thoughts. Leveraging text mining to provide best guesses regarding taxonomic names for illustrations, as well as recent experiments in identifying captions of illustrations in standardly formatted BHL journal articles are also inspiring.
Please stay tuned for more updates on my project in the next few weeks. In the meantime, I’d more than welcome talking about any of the topics above that interest you in the comments, or through email (firstname.lastname@example.org) or Twitter (@ari_rehbein)!