Rationale
1. Text
2. Images
3. Methods
1. Text
The Ecclesiastical Proust Archive does not directly answer questions about Proust but rather provides textual and visual occasions for experiencing À la recherche du temps perdu (In Search of Lost Time). Specifically, it is built on the recurring motif of the church in all its manifestations. This archive is perhaps an example of what Jerome McGann calls "textual works that are instruments of scientific knowledge" (19) and a document that makes text and image simultaneous to one another. Though McGann's remarks, as well as the D.G. Rossetti Archive, pertain to the oeuvre of a particular artist or artists, the Ecclesiastical Proust Archive pertains to a single work, albeit one that spans six volumes. In reorganizing Proust's text by means of a database, the user can perform a sort of scientific experiment by creating "sequential engagements with non-sequential forms of knowledge and experienceimmediate encounters with abstract or complexly mediated forms" (McGann 20). Though the search results are displayed in chronological order, they allow the reader to transcend large gaps in the narrative and behold at once a series of readings that are related, though separated by textual space. It allows "a much clearer and more capacious view of one particular class or 'order of things'" (McGann 20). However, this archive is one exclusively of internal relations, an interpretive archive of a single work.
The sole criterion for including a passage in the archive was whether it had any reference, direct or indirect, literal or figurative, to the idea of church, however that might be constituted. There are 336 church passages in the database. A valid passage can describe or imagine a physical church, can refer to a little clan as "the faithful" or assess their "orthodoxy", can compare a secular occasion to an ecclesiastical one, as in 2 2 1 613, where the narrator describes a dinner scene at Balbec as "an altar at which were celebrated the rites of the palate, and where in the hollows of the oyster-shells a few drops of lustral water had remained as in tiny holy-water stoups of stone," and so on.
Once a passage had been chosen, the next step was to decide where to begin and end its transcription. Every attempt was made to capture the entire moment of discourse in which the church participates. Some passages constitute a single sentence, others span several pages. For example, 1 1 1 174-176 is a two-page passage describing the full circumstances under which the narrator's family had to change their opinion of M. Legrandin. Since the defining moment occurred in front of the church at Combray, the entire two pages were transcribed in order to present the full discourse in which the church motif operates.
By contrast, the previous church passage, 1 1 1 171, is a single sentence comparing Françoise to kneeling figures in stained-glass church windows.
Perhaps the most important repeating element (in some cases non-repeating) is the group of associations. An association is a theme, concept, motif, important recurring detail, or some other structural feature present in the church passage itself. For example, a surprising number of churches in the novel are associated with meals or particular foods, so some of the more common associations are Food, Lunch, and Dinner. Many church passages describe one sense modality in terms of another, and therefore Synaesthesia is an association marking where this occurs. Another common association is Subject/Object, indicating where the ontological distinction between the experiencing subject and external object is considered by the narrator. Each passage has a number of associations connected to it in the database; they can be used in a search and viewed in the results. Further, the associations appear in the results as links: clicking on one initiates a new search that will recall all passages with which it is paired. This is a way of reorganizing the text and reading by chains of association. See the Instructions for more information on how this works.
In breaking up Proust's text into selected units of reading, and in inferring certain characteristics to be present in those units, this archive lends itself to the theoretical analysis of reading and narrative structure. At this point (7 May 2006) it is too early to say how studies based on this archive will take shape. Will users want to examine various narrative codes that pertain to the church motif? Will they want to map out the motif along the narrative arc of the novel? Will they ask exactly what a motif is? Will they want to inquire into the nature of absolute narrative continuity as opposed to gapped or selective continuity, and how these generate knowledge of the text? Will this archive successfully perform its intended task of generating knowledge of the text? Will it ultimately demonstrate a hierarchy of passages or associations within the narrative arc of the church motif, or in relation to the main arc of the novel? Can such arcs be said coherently to exist? Will it be useful to perform other similar studies within the Recherche and perhaps on other works by Proust or by other authors?
2. Images
The aspect of the archive that is specifically hypermediathe inclusion of imageslends itself both to the interpretation of the Recherche and to the study of theories of image and text. The images in the database are of two general kinds, representational and interpretive.
Representational images depict churches actually referred to (or re-named) in the Recherche. They are meant to supplement the highly visual descriptions of churches in the novel by displaying real churches that bear a relation to something specific in the text. This raises the question of whether such images can enhance the experience of reading Proust (or of any author). Do the images express the visual ambiance and 'spirit' of the novel? Do they further solidify that dimension of imaginary space? Or are they merely other representations that neither move the reader closer to nor further from the church signifier/signified?
Every effort was made to obtain images of all real churches mentioned in the novel. These include the cathedrals of Rheims, Amiens, Laon, Chartres, and more. When a fictional church has a real model, an image depicting the original is attached to the passage. This was done consistently. In that way, pictures of the Église Saint-Jacques at Illiers-Combray are always attached to passages about the Église Saint-Hilaire at Combray. Another of the more consistent pairings was with pictures of Chartres Cathedral and passages about Saint-André-des-Champs, since many of the architectural details were strikingly similarthe statues of saints in the porch and the twin steeples "tapering, scaly, chequered, honeycombed, yellowing and friable as two ears of wheat" (1 1 1 205).
Interpretive images are those aligned with passages in which churches are described but not specifically referred to. For example, where there is 'some church' imagined in a love fantasy, a picture is included that depicts the architectural features of the imagined church or a scene that evokes a certain aesthetic described by the narrator. The attempt was made, too, to enforce consistency between repeating types of passages and the types of images attached to them. Many of the passages in which the narrator describes a recurring fantasy of meeting a woman on the porch of a Gothic cathedral are attached to one of several overexposed photographs of the north and south portals of Chartres. These particular photos have an etherial, ghostlike effect that carries the mood of the fantasy passages though they aren't based on any particular model in the real world. This might or might not enhance the experience and understanding of the text and is a possible question to be taken up during the course of study. It is possible, too, to search the archive by properties of the images themselves, allowing the user further to investigate the relation of image and text.
For the most part, each image/passage pairing is unique, though there are a few repeated images where they fit several passages particularly well.
3. Methods
The Ecclesiastical Proust Archive is searchable by five different methodsText, Associations, Narrative Context, Image Properties, and Paginationwhich can be used in isolation or combination. Before searching the archive, please read the Instructions to learn how to use these methods.
The Text search enables the reader to recall and bring passages together on-screen by searching for words in the translated text. It is a standard text search with word, wildcard, phrase, and boolean phrase functionalities that allow a thorough and flexible interaction with the text.
The Associations search is a dropdown menu of all the associations contained in the archive. This function enables the reader to recall and organize passages by theme, concept, motif, and other structural features.
The Narrative Context search enables the reader to search passages by a menu of notes originally devised to help reconnect the decontextualized passages to the larger picture of the narrative. These notes usually indicate the location, the occasion, and the speaker, and tend to be unique to a single passage, though there are exceptions.
The Image Properties search is a dropdown menu of properties of the images in the archive. These can be the architectural style or part of the church depicted, the interior or exterior of the church, the medium (photograph, painting, drawing, etc.), whether it is in color or black and white, and so on.
The Pagination search enables the reader to recall a continuous series of passages as large or as small as desired. It can span multiple volumes or a single page.
The purposes and uses of this archive are essentially open-ended. Though organized under the somewhat narrow rubric of churches, it is a model that could, if successful, be applied to other interpretive rubrics such as gender, sexuality, memory, maternal relations, or family hierarchy. With much effort and time, the archive could even become a comprehensive database that records each sentence. That task could be performed in a different manner with qualitative analysis software, which interprets human language texts. However, the advantage of a human-translated archive is that it can embody nuances of meaning and connection that might not be picked up by the algorithm of a QA application. Whatever development this archive might undergo, it will be important to assess where and how it fails, as well as where it succeeds, in order to aid the creation of improved resources in the future.
McGann, Jerome J. "The Rationale of Hypertext." Electronic Text: Investigations in Method and Theory. Ed. Kathryn Sutherland. New York: Oxford UP, 1997. 19-46.