This is the class blog for “Text Mapping as Modelling,” taught by Øyvind Eide at the Digital Humanities Summer Institute in Victoria, Canada, from June 8–12, 2015.
Here are some titles and online references of special interest to this course:
Bodenhamer, David J., John Corrigan, and Trevor M. Harris. “Deep Mapping and the Spatial Humanities.” International Journal of Humanities and Arts Computing 7, no. 1–2 (2013): 170–75.
Drucker, Johanna. “Humanities Approaches to Graphical Display.” DHQ: Digital Humanities Quarterly 5, no. 1 (2011). Available online: http://www.digitalhumanities.org/dhq/vol/5/1/000091/000091.html
Eide, Øyvind. “Reading the Text, Walking the Terrain, Following the Map: Do We See the Same Landscape?” P. 194–205 in: Bode, K, and P. Arthur, eds. Advancing Digital Humanities: Research, Theory, Methods. Palgrave Macmillan, 2014.
Eide, Øyvind. Media Boundaries and Conceptual Modelling : Between Texts and Maps. Houndmills, Basingstoke, Hampshire: Palgrave Macmillan, Forthcoming 2015. Online presentation: http://tinyurl.com/n9ngckn
Fishwick, Paul. Philosophy of Modeling. Video presentation. Duration: 21:27. Available online: http://vimeo.com/26765714
Heat Moon, William Least. Prairyerth : (a Deep Map). Boston: Houghton Mifflin, 1991.
Kemp, Karen K. “Geographic Information Science and Spatial Analysis for the Humanities.” P. 31–57 in: Bodenhamer, David J., John Corrigan, and Trevor M. Harris, eds. The Spatial Humanities : GIS and the Future of Humanities Scholarship. Bloomington: Indiana University Press, 2010.
Monmonier, Mark. How to Lie with Maps. 2 ed. Chicago: University of Chicago Press, 1996.
Nowviskie, Bethany. “ ‘Inventing the Map’ In the Digital Humanities: A Young Lady’s Primer.” Poetess Archive Journal 2, no. 1 (2010). Available online: http://journals.tdl.org/paj/index.php/paj/article/view/11
Nowviskie, Bethany. neatline & visualization as interpretation. Blog post. Available online: http://nowviskie.org/2014/neatline-and-visualization-as-interpretation/
Wood, Denis, John Fels, and John Krygier. Rethinking the Power of Maps. New York: Guilford Press, 2010.
The first paragraph of Erich Kästner’s Fabian (1931) consists of a list of the headlines the title character notices while skimming the evening newspapers in a café. This prominence of world and local events through the lens of the daily press emphasizes the spatio-temporal and political situatedness of this “big city novel” (Großstadtroman). The first lines of any novel warrant close attention. When those lines highlight current events in a book explicitly time-sensitive, an investigation into the specificities of the headlines mentioned could be very revealing indeed. This certainly justifies research into each headline. The question in the context of our course, however, is whether digital mapping can be a useful tool to analyze and/or present the data unearthed by research.
After plotting the geographical and chronological sources of the news items in the headlines on a Neatline map, I remain unconvinced about the necessity or utility of digital mapping for the analysis of historical research in the case of this novel. For presenting the data, however, Neatline was an excellent tool. Several conclusions become visibly evident upon viewing the map and timeline of the events mentioned in the headlines. These results effect an interpretation of the time of the novel and the geographical perspective of its main character. In both cases, the research and analysis were possible without digital mapping, but Neatline makes the conclusions visually evident and compelling.
Time: The gesture of beginning a novel with newspaper headlines works to tie the fictional actions to a specific date in history. The very first headline, for instance, references the spectacular crash of the R-101 airship on its maiden overseas voyage (a kind of Titanic of the air). This event took place on October 5, 1930. It would hence be reasonable to assume that Fabian is sitting in the café on October 6, a Monday. Another headline, however, probably refers to the immediate aftermath of the parliamentary elections in Saxony, which took place on June 22, 1930. All the other identifiable headlines take place between June and October, 1930. It quickly becomes evident, therefore, that the fictional time of the novel is not a specific historical date. Instead, the smattering of news events from the summer and fall of 1930 evoke an imagined and impossible ‘everyday’—an abstract generalized day in the very concrete and particular historical year.
Place: In marking these points on a map, two interesting features become evident. First, the general distribution of points cluster densely around Berlin and then fan out in increasing dispersion. As might be expected from a random sampling of newspaper headlines anywhere, a larger number of stories deal with local events. The locations of reported incidents in Fabian unsurprisingly concentrate around Berlin (1 definite, 6 likely). The further one gets from this center, the sparser the locations become: 4 in German-speaking lands; 2 in other European countries; 2 in North America; 1 in South America. Second, the map makes clear what places are not mentioned in the headlines. Asia, Africa, and the Middle East are large expanses empty of all dots. The map thus reveals the occidento-centrism of the Berlin press and/or of Fabian’s attention.
In 8 CE, the Roman poet Ovid was sent into exile by the emperor Augustus. He traveled from Rome to a small city on the Black Sea coast, Tomis (modern-day Constanza, Romania). He composed and sent back to Rome many poems in exile, and the first book of these, entitled Tristia (“Laments”), gives an account of his journey. The narrative of this journey, however, is non-linear and incomplete. Consecutive poems jump from one setting to another (e.g., from a storm at sea to a scene of his departure from Rome and back), and large segments of his trip (including his arrival in his new home) are never mentioned.
In this project, I undertook to represent the travel-events described in these eleven poems in three different ways in order to see how each new visualization method enhanced and manipulated our understanding of the text.
Geographical Reconstruction Map
The first map I made was a reconstruction of Ovid’s historical journey, using as points geographical references made in the Tristia (marked in green) and scholarly speculations (yellow, orange) that fill in some of the narrative gaps. I included a (highly speculative) timeline derived from the one temporal reference Ovid gives in the book (that he was sailing in the Adriatic Sea in December).
This map allows us to imagine how the historical Ovid traveled through space and time to get to Tomis. While useful in this historical sense, it has the major drawback of conflating Ovid-the-historical-person and Ovid-the-character-of-his-poems. It assumes that the texts can and should be mined for biographical data, and also that recreating a historical journey will help us understand the narrative of the Tristia. I believe that both of these assumptions are problematic and deserve much closer scrutiny and discussion, but the map format does not admit their problems and inhibits any sort of discussion. Since this map, in laying these points out along a timeline, ignores the non-linearity that makes up such a remarkable feature of the text, I felt impelled to create a map that could reflect the journey that the reader takes while reading the book of poems. This led to my development of the following…
In my opinion, this was the least successful map, although I was most excited about it to start. I realized after a while that a geographical map (the same used in the reconstruction) was not well suited to representing this information. So I created a more conceptual, schematized “map” that represented the major areas that Ovid describes in Tristia 1: Italy (and Rome), the Sea, Greece and Thrace (as stopping points on his sea voyage), and the region he is headed to (often referred to as Scythia and associated with the edge of the world). This “map” helped address the problem that many of the poems in Tristia 1 do not specify a particular place of composition. The majority of them take the form of letters written to recipients in Rome (so their destination — and the fact that the destination and the origin are far removed — is far more important than the place of composition).
Ultimately, this way of representing the narratology of the book made me realize that it is the space in between that makes up the crux of book’s geography.
This mapping strategy has the most exciting potential, and could be expanded to examine the other books of the Tristia. Instead of trying to deduce a route or a poem’s setting (relying on data points that the author has made insufficient), this map represents geographical references that the poet makes over the course of the book using a contemporary map. The background image is a modern map based on the text of Strabo’s Geography (nearly contemporary with the composition of the Tristia). Red items represent places from which Ovid says he is leaving; purple items represent places he says he passes through; blue items represent regions he says he is going to. In this, the Ancient Map (unlike the Reconstruction Map) focuses on Ovid-the-poet’s own statements about geography. As the map shows, the regions are broad. For the most part, unlike the Reconstruction Map, they must be represented by polygons rather than discrete points. This emphasizes the fact that the poet is using broad, general, unspecific geographical terms. He also uses different (and conflicting) terms to refer to the same place (e.g. “Scythia” and “Sauromatae” are not the same, although they are proximal).
This map raises the question: how beneficial is geographical specificity? When and why does an author rely on non-specific geographical references? Plotting points on a map modifies the reader’s experience and makes all places equally accessible, but this does not reflect real reading experience. For all readers, some places are more familiar than others, and will conjure more specific images and associations; conversely, there are places that have an air of exoticism around them, that are less or unfamiliar and which the reader does not have real knowledge of. Using a map like the Reconstruction Map imposes a form of geographical knowledge and accuracy otherwise inaccessible for some/many/most readers. The reader will no doubt still provide mental images as s/he reads, likely an amalgamation of culturally-constructed stereotypes. An author may be drawing on this. I would argue that Ovid is doing exactly that when he refers to Scythia. He doesn’t expect his audience to have a clear conception of the geography; vagueness and exoticism are beneficial, for he can exploit the cultural associations (barbarism, coldness) that are not based in “factual” experience.
Jayne Eyre is a classic novel, written by Charlotte Bronte and published in 1847. It is about the life events of an orphan girl, set in the Yorkshire UK area, where the author lived. For this exercise I wanted to map the novel’s five major locations and other fictitious towns and places mentioned onto an actual historic landscape of the area. Bronte would likely have used her home location as the novel’s setting. Ideally I would have created a basemap using another mapping program such as ArcGIS, where I would create the towns, places and landscape using the details provided in the novel. This would require a good knowledge and close reading of the novel; as far as I know few details are provided of actual locations, distances between places, time (years), etc. Not finding a suitable basemap in Neatline or online, I found a pretty watercolour image which depicted the five residences to use as a basemap. It is not likely accurate as far as building details or distances between homes. The five residences are: Gateshead Hall, Lowood Institution, Thornfield Hall, The Moor House, and Ferndean Manor.
Next I found online images of each of the residences (most likely taken from movie settings) which I added to/embedded in the basemap, and added some explanatory text of events. This exercise simply provides an overview of the main residences and stages of Jane’s life, with some images to suggest what the residences looked like. It accompanies other exhibits on interior spaces and actual place names mentioned in the novel.
Erich Kästner’s Fabian (1931) opens in a café, in which the main figure sits while reading the evening edition of newspapers. Even though little description is provided for this place, it constitutes the enclosed starting point of the character’s later wanderings through Berlin. After leaving the café, Fabian goes to Frau Sommer, in the Schlüterstraße 23. Frau Sommer turns out to be the manager of a swinger club, the second important inner space of the novel. The narrator describes this space in greater details than the café and the reader learns, for example, that the office of Frau Sommer is located behind this curtain, which gives access to the other rooms: one in which people play bridge and another one in which people are dancing. A grammophon is playing in the background, “[i]irgendwo (12). Fabian would like to explore the other rooms of the club, but he never gets to do that because a blonde lady invites him to follow her to her place. The interior space of the cab in which the two are driving, even if not described, deserves to be included in the map. In the car, rolling down the streets of Berlin, the lady covers Fabian with love bites until they get to her place. Interestingly, there is no mention of the address at which the apartment of the lady is located; it is referred just as “eine Adresse” (14). The interior of the apartment is not described either, except for the bedroom. The reader is given details about the bed (“niedrig und breit”), the lamp (“gab indirektes Licht”), the wall (“mit Spiegelglas bespannt”), and the window (“[v]ergittert war es nicht”) (19).
The goal of my project was to map the interior settings described in the first pages of the novel. While the word “interior” may refer both to places located inside buildings and to imagined, internalized spaces, my visualization captures only the first meaning – in part, also because of the difficulty of mapping psychological spaces with a tool like Neatline.
Neatline allowed me to develop my spatial analysis in different steps. First of all, I was able to use a historical map of 1929 Berlin, the time in which the novel is set, adding a historical layer to the mapping. While one could do this by accessing a print out of a map, the immediate access and use of the historical material was definitely an advantage in getting our project off the ground. With the basic function of dropping points, I could first identify in the map the address of the internal spaces, when indicated. This helped me to visualize where the events happened in Berlin and, at the same time, to realize how some addresses were provided in full while others were completely omitted. The polygon function made it possible to signal a broader geographical area, when the precise location was not provided in the text. However, as it emerged in the follow-up discussion of the project, a polygon, which per definition has borders, still provides a specific collocation for something that the narrator left intentionally unspecified. One could, of course, use a text annotation to make the viewer aware of this ambiguity, but once something is put on a map it is given a specific position – or at least this is what the viewer perceives.
To provide some guidance in the clicking, I decided to use the plug-in Waypoints, which allowed me to organize the spaces as they chronologically appear in the novel, thus mirroring the narrative development. Waypoints is a very nice function because it organizes the information on the map in a clear, readable way, avoiding, at the same time, to have to use the Neatline timeline. I was happy with this choice for two reasons: first, a feature of the novel is its temporal unspecificity, at least for some of Fabian’s movements; second, the Neatline timeline requires very specific temporal data in order to function and display properly, which would enforce a more precise time frame on the text. In an expanded version of this project, I imagine my part to be embedded in the broader mapping of “Fabian’s movement through Berlin,” which already includes a timeline for the different movements, making a second timeline redundant. With this regard, it would be helpful if Neatline actually allowed to display different maps that belong to the same project within one exhibition.
In a next step, I drew maps of two interior spaces, Frau Sommer’s club and Frau Moll’s apartment, based on the description given in the text. By doing this, I transitioned from mapping the text to creating a map for the text. I made these drawings using a home styling program and I had to imagine some of the details that were not directly mentioned in the text.
The annotations, quotations, images and sounds I added to describe the points aimed at providing more contextual information for the specific locations, while maintaining a direct connection to the primary text. I think this is a function that, in a teaching situation, can be used to engage students, either by asking them to annotate the map themselves or to reflect on the annotations.
Reflection and Open Questions
But once all this information is added to the map, what does this bring to my analysis? What does the map offer that a textual representation would not? I think the mapping itself can be really helpful in prompting a meticulous reading of the text, since every relevant detail provided in the text needs to be translated onto the map. Even though one could collect this information also using pencil and paper, I think the possibility of editing the historical map, of resizing it to mirror how the text changes its focus, while at the same time creating an overview of the whole text, is an advantage of the mapping process. I would say that mapping may not be a necessary tool to implement in the classroom or in the research, but it may be helpful depending on the task. If one were to complete a socio-economic reading of Berlin in 1930 based on the descriptions provided in Fabian, for example, the map would become an essential tool to understand where different places are located and in which parts of the city different activities takes place. One could even reflect on those spaces that remain untouched by the narration, in relation to the places that are mentioned. Maps alone are probably not leading to a total new understanding of the text, but the hybrid nature of geocommunication can add a new layer of meaning to analysis.
The visualization of the spaces on the map, both in relation to the rest of the city and within the buildings, has the advantage of offering a clear and clean way of representing the text, translating words into images. Further, it provides an immediate context and orientation even to readers that may not be familiar at all with the setting of the novel. The possibility of adding annotations that complete the geo-spatial representation of the spaces also makes it possible to complicate the visualization and direct the viewer’s attention to specific details. And yet, while all these functions allow one to visualize the interior spaces described in the novel, the mapping is not able to capture the vagueness of some text passages. Indeed, the visualization of elements works in a counter direction to the openness expressed in the text. So, after spending the last week mapping the text, the following questions remain: How can one suggest the literary/narrative uncertainty which characterizes the spatial descriptions in Fabian or any other literary work? To what extent do some mapping functions, such as the zooming for example, suggest a hierarchy of places/spaces that may not be actually described in the novel? How does one need to provide guidance for the user/viewer of the visualization in ways that may be different from the ones suggested in the text? And finally, if one were to attempt to represent spaces of psychological interiority, what possibilities would Neatline allow for?
Having just presented projects to the group, it seems as though the majority of my progress this week is with reference to what I didn’t achieve, rather than to what I did. To explain. First, although my technical confidence has increased, I have not worked out some ‘tweaks’ which would have created a far more professional and visually fluent Neatline. Second, the Neatline exhibits I made provide visual interest; however, they do not necessarily provide digital insight. One of the issues I have become acutely aware of, is that the full potential of Neatline is only manifested when it is used as a tool for interpretation, not just visualisation; my first Neatline of the Red Room in Jane Eyre, was exactly that, a visualisation. The second mapping, of Mr. Rochester’s room, began to show something more interesting; by creating a map in plan mode, rather than elevation as for the Red Room, the movements of the characters revealed an intensity and frenzy which was not so apparent in Bronte’s novel. In the narrative, there were reflective moments which dissipated the frenzy. This second map focused my mind on the significance of map modelling ‘of’ and map modelling ‘for’; achieving the latter is when the full potential of Neatline is evident but requires a starting focus on the research question rather than the text.
Ashli, Theresa, and I adapted Ernest Hemingway’s “The Killers” for our neatline project. The short story is set primarily in a saloon turned lunch-room and centers on two strangers who intimidate and contain the waiter, cook, and patron who are in the space when they arrive. They wait for a specific customer, Ole Andreson, whom they’ve been sent to kill. Several customers come in and are sent away, and then the two strangers finally leave. The patron Nick, who had been held at gunpoint and tied up most of the narrative, goes to warn Andreson. Andreson has been in his room in a rooming house this whole time, where he’s been immobile, awaiting a fate he perceives to be inevitable. Nick speculates that the trouble must have arisen from Chicago, and considers leaving town.
We used our map to attempt to faithfully adapt the short story’s linear plot time and detailed descriptions of the characters’ movement in space. We used icons of color-coded hats to track the movement of the characters through a visual representation of the lunchroom and the route to Andreson’s rooming house. We tied these movements to the progression of time established in the short story, using Neatline’s “simile” plugin. We also associated each point in space with relevant dialogue from the short story and some pictures from Scribner’s Magazine, which the story originally appeared in.
In our adaptation of the setting to the map format, we made decisions about how to visually represent things that were not detailed in words. The lunch-counter where most of the dialogue occurs, for example, may look very different to different people, based on their experiences (diners they’ve been to, images of lunch-counters from the 1920s they’ve seen images of, etc). In our iteration, the lunch-counter looks a specific way, with black and white linoleum. We therefore added to the text and fixed elements of it in ways that a print text does not. We also did so in a simplifying and modular way, however. We represent the cook Sam through a chef’s hat icon rather than the much more detailed representation of him in the illustration that accompanies the story.
With a linear narrative, a narrow focus on one place at a time, concrete language, and broadly drawn characters, the short story “The Killers” does not provide much beyond the concrete, material and realist. Place exists and is built through asides, mentions, and so on, rather than through thick description. As such, focusing on placing the dialogue as points in the space seems to put the emphasis on a part of the story that is less interesting, and impoverishes the rich dialogue. This particular story may not be especially suited to a concrete and straightforward map. Had we mapped power structures in the story, for example, or a Nick’s point of view of the story, we might have had a complementary resource. If the narrative were more layered or otherwise nonlinear, if the language were more figurative, if the characters were more complex, and so on, then a simplifying and concrete visual model might also be more interesting. Something like Murakami’s 1Q84 or David Mitchell’s Cloud Atlas might provide fascinating maps.
In terms of our technical skills, we only used the tip of the tool’s possibilities. We were unable to fluently use all of Neatline’s capabilities and, should we make a second draft of the project, we would make several changes. First of all, we would do something other than a literal adaptation of the story. Even if we were to have kept the linear map, we would have been able to tell the story better with waypoints than with the timeline. We would also have liked to create layers for each character to better represent their movements in the space.