Textual Iconography of the Quixote: A Data Model for Extending the Single - Faceted Pictorial Space into a Poly - Faceted Semantic Web

Since its initiation in 1995, the Cervantes Project has focused on creating a comprehensive on-line
resource centered on the iconic Hispanic author Miguel de Cervantes (1547-1616). The project, a collaboration of researchers in Hispanic Studies and Computer Science,
provides bibliographical, biographical, and textual
materials, including an Electronic Variorum Edition based on first editions of Cervantes’ best known work, Don Quixote. Recent additions include an exhibit of bookplates (ex libris) inspired by the Quixote. Soon to be released are a significant collection of illustrations
associated with key Quixote editions (the textual
iconography) and a presentation of the musical aspects and influences of Cervantes’ works. These latter two
collections will be discussed further in this paper.
As our work with the Cervantes Project has evolved, we have become increasingly aware of the need to
develop rich interlinkages between the panoply of resources
collected for and produced by this project [3]. Within the broader scope of the project as a whole, our textual iconography collection offers a fertile ground for exploring the needs of large, interdisciplinary collections, and the
techniques for integrating the resources they contain.
The textual iconography collection (Figure 1) is intended
to facilitate a more complete understanding of the iconic
transformations of the Quixote [5] throughout history
by assembling a digital collection of more than 8,000
illustrations taken from over 400 of the most significant
editions of this seminal book [15]. In the process of
assembling and working with this collection, we have
found traditional approaches to presenting image-based
collections in both print and digital media to be lacking
[12]. While print has long been the dominant media
of scholarly communication, it does not scale well to
projects of this scope [8]. Digital image collections, on
the other hand, are often focused on presenting large
numbers of images and associated metadata but lack
the scholarly commentary and integration with other
textual resources required for scholarly investigations.
Moreover, we have found a need to adopt a rich model
of hypertext, based on automatically-generated links
and multiply-rooted to reflect multiple interpretations of
simultaneous significance.
Figure 1: The textual iconography’s browsing interface allows viewing of the collection
and its metadata. The collection’s editors also have access to maintenance tools. We have aggressively pursued the development of this collection 2 with the intention that it will enable new forms of textual, visual and critical analysis. In particular, it enables scholarly research in two major directions that remain poorly understood despite the incredible amount of scholarly attention devoted to the Quixote. First, the textual iconography of the Quixote is a key resource to help literary scholars better understand its reception and interpretation throughout history—the illustrations acting as a “hand-mirror 3 ” that allows us to see how successive generations of readers, including our own, have literally painted themselves into Cervantes’ story [7]. Second, this collection is an invaluable dataset for art historians enabling them to better explore the tools and techniques employed in the often neglected field of book illustration 4. The illustrations found in the pages of the Quixote trace the evolution of graphic art in modern
printed works, from the first wood cuts of early 17th
century, to the copper engravings, etchings aquatints of the middle and late 17th century and 18th century, to the xilographs and lithographs of the 19th century, and finally the mechanical techniques of the 20th century including
offset printing [10,13,14]. These areas remain poorly
understood, not simply because of a lack of attention, but in large part because traditional approaches do not
adequately support investigations of this nature.
The value of the textual iconography collection is not limited to the scholarly community. Its canonical status
in world literature courses, its iconic nature in Hispanic culture, and the renewed interest awaked by the recent celebrations surrounding the 400th anniversary of its publication, ensure a continued popular interest in the Quixote. The illustrations of the adventures continue to captivate the popular imagination and help to bring
the book alive to millions of readers. By making the
collection available to the public in digital format, we are able to radically increase the resources available to students of the Quixote from all walks of life. Our goal, however, is not simply to provide a large pile of
pictures for students to peruse, but to further enrich their
understanding of both the Quixote and of art history by supplementing the illustrations with commentary about the significance of the images, their relationship to the text, and the artistic achievement of the illustrators and engravers who created them.
To adequately support both the scholarly and popular
communities that will access the collection, it is not
sufficient to merely provide an online catalogue of the illustrations, regardless of the completeness of the
collection or the detail of the metadata. Instead, we
have identified three levels of information required to successfully present these materials:
1. Descriptive metadata: The first level of information
includes factual descriptions of the items it describes.
For our collection, descriptive metadata is provided
for both individual illustrations and the editions in which those illustrations were published. This
information includes the title of the image, its physical size and the size of the page it is printed on, the name of the artist and/or engraver, the style of the image, and the printing technique used.
2. Scholarly commentary: The second level goes beyond simple descriptions of the items in the
collection to provide a critical assessment of those items, their narrative context, and their hermeneutic and aesthetic significance. Within our collection we are developing biographical commentary about artists and engravers and technical and artistic commentary about each individual image.
3. Hypertextual connections: The third level consists
of the information needed to develop dense
hypertextual structures and the tools to effectively navigate them.
A key challenge in constructing collections of this type is developing appropriate strategies for interconnecting the resources being continuously added to the collection.
While most traditional editorial approaches require
editors to carefully edit each resource by hand, this is beyond the scope of our current resources and would greatly limit the size of our collection, the speed with which it could be made available, and the degree of
interlinkages that could be provided.
Accordingly, we are using two strategies for automatically
interlinking digital resources. One approach relies
heavily on the metadata associated with each digital image.
We use this metadata information to automatically
discover relationships between existing and new resources
and then to generate navigational structures based on those relationships. A prototype of this system has been employed internally for use in a project centered on the music in Cervantes’ works [9] (Figure 2). One key
application of this approach allows us to integrate multiple resources, including illustrations, with the textual and narrative structure of the Quixote [2]. To accomplish this we have developed a formal taxonomy of the narrative and thematic elements of the Quixote, which is then used in cataloging each image and can be used for the associated commentary. This allows us to codify the connections
between the illustrations and commentary without requiring
a particular version of the text. The second approach
to interlinking resources we have developed involves elucidating the internal structure of the textual resources in the collection and using that structure to automatically generate navigational links and to inform visualizations
[1]; an approach that generalizes Crane’s [4]. In this
context, we have developed a toolkit to implement this
approach for a collection of historical documents pertaining
to Cervantes and his family.
Figure 2: The display of resources pertaining to Cervantes’ works is automatically constructed from fragmentary entries. Structured links are generated by the system to allow access to a range of resources associated with key phrases. The hypertextual collections created for our purposes
create separately rooted structures over a common set of interlinked elements. For example, in the music collection
referred to above, natural collection roots include the
compositions, the composers of the pieces, the instruments
associated with the pieces, and the texts that refer to the pieces. Each rooted collection includes information
specific to the collection (e.g., biographies of the composers
and musical scores are associated with their respective collections), but ultimately collections are cross-linked
(e.g., between compositions and composers, between
instruments and compositions). The texts provide a unifying
framework that draws the other components together, yet the texts themselves represent a distinct collection. The structures in the textual iconographic collection show
similar relationships, with the added complexity implicit in the three-level categorization of collection information.
In conclusion, by adding detailed scholarly commentary
in addition to descriptive metadata and to providing
sophisticated tools for automatically building a hypertextual
structure into the collection, we are able to provide a new resource that better meets the research needs of scholars
and to assist the general public in understanding and
appreciating the significance both of the Quixote itself, as well as its reception and visual history. Our approach allows the readers of scholarly commentaries more direct access to the primary source materials used to develop and support those commentaries 5 [6,11]. Conversely, it allows individuals focusing on the primary materials
access to secondary scholarly works to better understand
a variety of reading perspectives and to explore and formulate their own interpretations of these unique
materials. Moreover, the nature of the collection and its information relationships has required the addressing of canonical hypertextual structuring issues. Taken as a whole, our collection and hypertextual archive opens previously unavailable opportunities for scholarly study of the Quixote and its unique literary, cultural and iconic status.
Footnotes
1. http://cervantes.tamu.edu/
2. In 2001, the Texas A&M University Cushing Memorial Library began an ongoing effort to assemble a rare book collection comprised of all significant illustrated editions of the Quixote. Currently this collection contains 413 editions in 15 languages: Don Quixote
Illustrated: An Exhibit in Celebration of the 4th
Centenary of the Quixote, 1605-2005. Texas A&M University; Austin: Wind River Press, 2005.
3. John Harthan points out that “book illustration is like a hand-mirror in which one can see reflected great historical events, social changes and the movement of ideas down the centuries” (Harthan, 1981).
4. Indeed, Harthan states that “a history of modern book illustration could almost be written in terms of this perennially popular classic alone” (Harthan, 1981).
5. This echoes our previous work in developing an electronic variorum hyperedition that gives readers access to the primary textual sources used to reach editorial conclusions, thus allowing readers to form their own opinions about which of several variants more likely constitutes the correct text of a disputed passage. http://www.csdl.tamu.edu/cervantes/V2/variorum/index.htm
References
1. Audenaert, Michael Neal. 2005. “Feature
Identification Framework and Applications.” M.S. Thesis, Texas A&M University.
2. Audenaert, N., Furuta, R., Urbina, E., Deng, J., Monroy, C., Sáenz, R., and Careaga, D. (2005)
Integrating Collections at the Cervantes Project. Proceedings of the Joint Conference on Digital
Libraries, JCDL’05, Denver. 287-288.
3. Audenauert, N., Furuta, R., Urbina, E., Deng, J., Monroy, C., Sáenz, R., and Careaga, D. (2005)
Integrating Diverse Research on a Digital Library Focused on a Single Author (Cervantes). Lecture Notes in Computer Science: Research and Advanced Technology for Digital Libraries: 9th ECDL, Vienna. 151-161.
4. Crane, G., Wulfman, C. E., and Smith, D. A. (2001) “Building a Hypertextual Digital Library in the
Humanities: a Case Study on London. Joint
Conference on Digital Libraries, JCDL01, Roanoke. 426-434. 5. Cervantes, Miguel de. Don Quijote de la Mancha. English translation by Edith Grossman (New York: HarperCollins, 2003).
6. Kochumman, R., Monroy, C., Furuta, R., Goenka,
A., Urbina, E., and Melgoza, E. (2001) Towards
an Electronic Variorum Edition of Don Quixote.
Proceedings of the First ACM/IEEE-CS Joint
Conference on Digital Libraries. 444-445.
7. Harthan, J. (1981) The History of the Illustrated Book: the Western Tradition. New York: Thames and Hudson.
8. McGann, J. (1997) The Rationale of Hypertext. In Sutherland, K. (ed.): Electronic Text: Investigations
in Method and Theory. Oxford UP, New York.
19-46.
9. Pastor, J. J. Música y literatura: la senda retórica. Hacia una nueva consideración de la música en
Cervantes.(2005) Doctoral Dissertation, Universidad de Castilla-La Mancha.
10. Urbina, E., et al. (2004) “Iconografía textual del Quijote: repaso y nueva aproximación de cara al IV centenario.” Le mappe nascoste di Cervantes. Actas
Coloquio Internacional de la Associazione Cervantina di Venecia (Venice, 2003). Carlos Romero, ed. Treviso: Edizioni Santi Quaranta. 103-114.
11. Urbina, E., General Editor and Director. Electronic variorum edition of the Quixote. Cervantes Project, Texas A&M University. http://cervantes.tamu.edu/V2/variorum/index.htm (accessed November 15, 2005).
12. Urbina, E., Furuta, R., and Smith, S. E. (2005) Visual Knowledge: Textual Iconography of the Quixote, a Hypertextual Archive. Proceedings
Abstracts ACH/ALLC 2005, Association for Computers in the Humanities. University of Victoria (Canada). http://web.uvic.ca/hrd/achallc2005/abstracts.htm (accessed November 15, 2005).
13. Urbina, E. (febrero-marzo 2005) Don Quijote, libro ilustrado. Contrastes (Valencia) 38, Special issue: “Quijote intesdisciplinar”. 37-41.
14. Urbina, E. (2005) Iconografía textual e historia
visual del Quijote. Cervantes y el pensamiento
moderno. José Luis González Quirós and José María Paz Gago, eds. Madrid: Sociedad Estatal de
Conmemoraciones Culturales (in press).
15. Urbina, E. (2005) Visual Knowledge: Textual
Iconography of the Quixote. Don Quixote Illustrated:
Textual Images and Visual Readings. Eduardo Urbina and Jesús G. Maestro, eds. (Biblioteca Cervantes 2). Pontevedra: Mirabel Editorial. 15-38.