Using Google's Cloud Platform, this project will develop a technological infrastructure for the visualization and distribution of data associated with “How Poetry Mattered in 1920s Korea,” one of two current book projects. With Alex Amies, Google Cloud Computing.
Classic Literature as Sculpture
So much of work that manifests our literary classics is hidden from view, especially now that we read literature using computers so much of the time. "Reprinting 'Azaleas'" attempts to give a shape to some of the computational procedures that materialize our favorite literature. Here we have created three dimensional models out of the Unicode (see this primer if unicode is unfamiliar) used to present that well-known Korean poem "Azaleas" by Kim So-wŏl on the web, in word-processing programs, and even in printed books. Please see the article by Wayne de Fremery and Sanghun Kim in "Reprinting 'Azaleas' — Meditations on Volume and Volumes" in AZALEA 9 (2015): 366–402.
We have taken the idea of "modeling texts" rather literally, creating models of texts that can be printed as models using 3D printers. Pictures of the models that we have printed appear in the article mentioned above, "Reprinting 'Azaleas' — Meditations on Volume and Volumes" in AZALEA 9 (2015): forthcoming. Images of models that we have printed will be available here once the article is published. In the meantime, please enjoy versions of the models we have made available online.
The serious aim of our quixotic critical move is to shed light on the socio-technological procedures that are sending our literary classics into the future. This topic has lacked a critical space in discussions of Korean literature. By lending "volume" to some of these procedures we hope to enable literary and historical insight while presenting artistic opportunities to reimagine what literature can be.
Classic Texts as Traversable Terrain
Here we have modeled the Unicode values associated with "Azaleas" to resemble a landscape and made the model (below) the terrain of a first-person adventure game. An image of the cover of the Hansŏng Tosŏ issue of Kim So-wŏl’s 1925 Chindallaekkot "textures" the mountains and valleys.
Text Mining and Search without OCR (optical character recognition)
Mo文oN is prototpe software that was initially conceived and developed as a new research and learning environment for Korean literature and culture. We quickly realized that the environment could productively benefit the study of any literary or cultural tradition. Mo文oN's current iteration includes a variety of patent-pending text mining and analysis algorithms, including what we call "Cut & Search." Cut&Search enables document images to be searched and mined without the need for OCR (optical character recognition) or other methods of encoding. This enables paleaographers, bibliographers, literary scholars, and historians to search, mine, and compare a variety of texts and textual witnesses.
Mo文oN also enables the modeling of encoded texts, which often leads to insight. Please see Wayne de Fremery and Sanghun Kim, "Remodeling Creation: The 1922 Issue of Kaebyok Magazine as graphs, maps, and (soon) trees," Journal of Korean Studies 20, no. 2 (Fall 2015): forthcoming.
Geographical Information Systems (GIS)
In addition, Mo文oN allows geospatial analysis of texts and textual information. To create the image below, a historical map of Seoul (1927) was associated with information about literary publishers and then "dropped" into the Mo文oN environment to reveal publisher locations in colonial Seoul.
Digital reproductions of cultural objects are routinely used as information and/or evidence in a wide variety of studies. Yet, many of the procedures and mechanisms that constitute digital objects are hidden from view. This article documents an experiment designed to reveal what is ordinarily concealed when digital reproductions of textual artifacts are enacted. We describe remodeling an encoded copy of the July 1922 issue of the important Korean colonial-era periodical Kaebyŏk (Creation) hosted on a website administrated by the National Institute of Korean History (NIKH). This was done by generating an alternate visual scheme for the Unicode values used as part of the computational procedures that present the July 1922 issue of Kaebyŏk on the NIKH website. We also document our attempts to account for the patterns that emerged when we remodeled the issue.
This creative engagement with the materials and computational elements of digital texts revealed itself to be a fecund form of literary, textual, and historical critique. Indeed, our experiment generated a number of domain-specific hypotheses about genre, translation, and compositional practice in colonial Korea while raising important questions about studies of culture that bracket either the material or discursive systems of cultural objects from view. A narrow focus on the material-discursive details of a single issue of one periodical brought expansive, largely unexplored fields of investigation into view
The Adan collection houses approximately 90,000 rare volumes dating from the sixteenth to the twentieth century. Roughly 10,000 premodern titles (chong), which are frequently many volumes long, are contained in the Adan archive. The collection of modern materials (late nineteenth / early twentieth century) at the library consists of approximately 41,000 monograph and periodical titles. In addition, the papers and personal effects of a number of important twentieth century authors are archived at Adan, along with eighteen texts that have been declared "national cultural assets" by the Cultural Heritage Administration of South Korea.
The Korea Text Initiative at the Cambridge Institute for the Study of Korea, in collaboration with members of the Electronic Cultural Atlas Initiative at UC Berkeley, is helping to make the collection at Adan available online.
Classic Literature as Immersive Environments
The Immersive Poetry Project explores the future of Korea’s cultural past—a future where Korea’s literary classics can be traversed like Korea’s many forests and mountains. Our aspiration is to deepen the discussion of vernacular Korean poetry by enabling engagements with Korean poetic texts that privilege image over discourse, if only temporarily. We attempt to see, quite literally, what Korean poems can be in order to deepen discussions of what they are or might mean. Importantly, our new means of presenting Korea’s literary heritage reveals that heritage in a way that anyone, Korean and non-Korean alike, can experience. Our poetic landscapes are generated by the beauty of the Korean language but, like Korea’s picturesque woodlands, the Korean language is not a prerequisite for enjoying a stroll through them.
Nature Never has been exhibited at 6th International Conference on Intelligent Technologies for Interactive Entertainment in Chicago (2014) and HCI Korea in Seoul (2014), as well as at Sogang University. For a more detailed description of the environment please see Wayne de Fremery and Jusub Kim, “Kim So-wŏl’s Chindallaekkot (Azaleas) as an Immersive Environment,” Acta Koreana 17, no. 1 (June 2014): 5–27. The article is available online here.
We have also expressed the book of poetry as a variety of virtual reality experiences and are currently conducting user studies to understand how students read classic poetic texts differently when they are articulated as printed facsimiles and scholarly editions, as well as digital texts and virtual reality experiences.
The image above is a link to an interactive bibliography that presents information associated with forty-five individual copies of vernacular books of Korean poetry published between 1921 and 1929. The data are from Wayne de Fremery, "How Poetry Mattered in 1920s Korea" (PhD diss., Harvard University, 2011). The technology we use to present this data was initially developed by Microsoft and utilized Microsoft's Silverlight plug-in. Here we use HTML5 and are workin to extending the capabilities of the original Silverlight plug-in to include GIS and network analysis of the data. Many thanks to Natasa Milic-Frayling at Microsoft Research Cambridge (UK) for helping to create the first version of this presentation, which is still available. Thank you to Howie Lan at UC Berkeley for his help developing HTML5 version.
"Cut & Search" in Mo文oN Text Mining and Search without OCR (optical character recognition)
Mo文oN is prototpe software that I am developing with Sanghun Kim. It was initially conceived and developed as a new research and learning environment for Korean literature and culture. However, we quickly realized that the environment could productively benefit the study of any literary or cultural tradition. Mo文oN's current iteration includes a variety of patent-pending text mining and analysis algorithms, including what we call "Cut & Search." Cut&Search enables document images to be searched and mined without the need for OCR (optical character recognition) or other methods of encoding. This enables paleaographers, bibliographers, literary scholars, and historians to search, mine, and compare a variety of texts and textual witnesses.