“Toward a Spatial Narrative of the 1935 Harlem Riot: Mapping and Storytelling after the Geospatial Turn,” New Approaches, Opportunities and Epistemological Implications of Mapping History Digitally: An International Workshop and Conference, German Historical Institute, October 20.

ABSTRACT

This lecture considers spatial narrative as a form of digital scholarship, and particularly as a means of building arguments and interpretations from the maps combining and displaying sources that have proliferated with the expansion of web-based mapping. It explores the process of shaping a narrative of the 1935 Harlem Riot based on Year of the Riot, a map developed as an extension of the award-winning Digital Harlem project. A riot offers a particularly rich example for exploring the possibility of spatial narrative: an event that, while beginning at an identifiable origin, spirals out in multiple directions and changes over time in ways that expose the limitations of the existing linear web-based story-map authoring platforms. I prototype to what extent I can present a narrative of the 1935 Harlem riot using Neatline.

TEXT

There is only one way to begin this talk: with a map.

It is an example of the kind of historical mapping that resulted from the rise of web mapping, from the geospatial turn. It’s a mashup, featuring point data plotted on a historical basemap overlaid on Google Maps.The map is produced on a new site called Year of the Riot: Digital Harlem, 1935, a redesigned extension of Digital Harlem, the site my collaborators at the University of Sydney and I developed more than a decade ago. It relies on a different database, Heurist, that allows faceted search and a timeline (and its still in development, so not yet public).

Like Digital Harlem, Year of the Riot is a research tool, combining and visualizing a range of different sources to allow researchers to look for spatial patterns. While it doesn’t offer the kind of tools for spatial analysis provided by ARC-GIS, the Year of the Riot does combine and organize a disparate collection of qualitative sources and allow a user to explore and juxtapose different lives, places and events to address humanities questions. It does not offer any interpretation, argument or narrative; it simply displays the data. This map of the events of the riots does include an argument in the sense that it is shaped by the choices we made about what to include in the database, and how to categorize and present that information. But that kind of argument is very different from providing a structured way into and through the information. Thanks to the timeline, the map shows what, where and when; you can use the timeline to move sequentially through the events on the map. But Year of the Riot does not address the connections between those events, does not offer the explanation of why they occurred and what they mean that is the stuff of historical narratives and arguments.

We are now at a point where we want more from mapping than a means of exploring sources. Digital historians are facing calls for digital history to make arguments, and are looking to expand their engagement with the discipline in order to expand the use of digital tools and to integrate them with conventional methods. We’re finding that many other audiences do not want to explore a map to find an argument or construct a narrative, or are not equipped to do so. Without a place to begin, and a pathway to follow, they spend little time engaging with digital maps.

Scholars working with ARC-GIS rather than web-mapping are arriving at the same point. For almost a decade, they have been conceptualizing humanities mapping, moving away from historical GIS, to spatial humanities, to deep maps, and finally to spatial narratives. That discussion has remained largely conceptual and methodological: as Trevor Harris argues, “Deep mapping, spatial storytelling, and spatial narratives are incomplete terms struggling to capture and imbue meaning to abstract thoughts of a more profound, insightful, reflexive, multimedic, perhaps quixotic representation of humanistic space than currently prevails.”¹ The focus of GIS on spatial connections, patterns and relationships, and difficulty in dealing with time, has frustrated efforts to use it to answer humanities questions concerned with process, structure, and event in the context of place. Deep maps are the first part of imagining something different: a map of place not space, a shift in perspective from an emphasis on measured space to the connotations, associations, and meanings that people attach to spaces. A map of place involves a heterogeneity of different kinds of sources, multiple viewpoints and competing perspectives, and spatial data of different degrees of ambiguity, uncertainty and imprecision. That plural character means deep maps must be explored; by design, they do not offer any kind of metanarrative.

However, deep maps are not conceived as an end in themselves, but as the basis for constructing a spatial narrative within which an argument is embedded. The concept of spatial narrative addresses the same need for maps to be more than something to be explored raised by calls for digital historians to make arguments. Deep maps provide the context for spatial narratives, which are specific pathways through the map designed to tell a story. Conceiving spatial narratives as pathways highlights their visual character; they appear on a map. In that form, they require a specified point of departure, a series of pre-selected locations (and evidence) and an endpoint. However, there is tension between the linearity inherent in the notion of a pathway and the ability to capture complexity that maps offer: a pathway forces a narrow focus, allowing only one location for each point in time, at odds with ability of a map to capture “the simultaneity of lived existence,” to integrate and visualize multiple events happening at the same time.²

A spatial narrative has to be more than a pathway if it is to tell a story. A sequence of points can readily convey what, where, and when; a narrative is concerned also to offer an explanation not just chronicle events. How much explanation visualization can provide, how much text needs to be included, and how the visual and textual are located and balanced, are key issues in designing spatial narratives. So, where the concept of spatial narrative as constructed from deep maps begins with/gives preeminence to the map, to the visual, Lincoln Mullen’s definition of a narrative map, by contrast, begins with the textual: “A narrative map tells a story plotted through space. The point of a narrative map is not to display data. Rather it is to provide an explicit visual counterpart to the implicit spatial underpinnings of a narrative or argument.”³

Proponents differ over whether spatial narratives should be part of a deep map, or be created apart from it. At present, there is no deep mapping platform available that incorporates the ability to build spatial narratives. YOR certainly does not; the closest it comes is allowing the creation of featured maps that can be accompanied by a box of descriptive text. However, there are a number of web-based authoring platforms that can be used to create spatial narratives using data from a deep map: StoryMap JS, ESRI’s Story Map, and Neatline. These platforms offer an opportunity to move discussion of the design of spatial narratives from the conceptual and methodological to the concrete.

StoryMap JS is a creation of the Knight Lab at Northwestern University, intended initially for journalists. It consists of a simple editor synched to Google Drive that allows you to create a sequence of slides, each split between header text, explanatory text and optional media, such as an image or embedded video, and a point on a map. A line connects each point to the next location in the sequence. A reader navigates by clicking on an arrow to advance the slide or by clicking on a point on the map, which then moves dynamically to the next location and slide. The interface also offers a map overview, showing all the points.

StoryMap JS allows only points, and is strictly linear, permitting only a sequential arrangement and a view oriented to each slide. The animated movement between slides enhances the sense of being on a pathway through the map. Being able to incorporate images and video offers the possibility of balancing the aerial view of the map with a perspective on the ground. However, while the split screen of each slide balances text and map, the limits to what you do on the map make this a platform that delivers a story plotted through space rather than a map driven narrative. StoryMap JS is a platform for adding spatial elements to a narrative more than constructing a narrative drawing on what a map offers.

Vincent Brown’s custom-built spatial narrative, Slave Revolt in Jamaica, 1760-1761, is a pathway through a map that relies more on visual features, in which, he argues, “the interpretive focus of [the] story emerges from its visual design.”⁴ Slave Revolt employs a basemap of topography, symbols of different colors to differentiate the fighting units, and icons for conspiracies, clashes and judicial executions. A fading tracer line tracks movement. The map can be animated to show “the unfolding of historical processes,” as events driving the movement and suggesting causes and effects, or manually clicked through using arrows. ⁵ Pop-ups linked to points on the map include just a few words of information; if the map is animated, clicking a popup pauses the map until you move your cursor. The map is supplemented by a narrow panel of text, which displays excerpts from 18^th century documents, not an argument or narrative.

Brown argues that by “Tracing [combatants] locations over time, it is possible to discern some of their strategic aims and to observe the tactical dynamics of slave insurrection and counter-revolt.”⁶ However, I certainly would not have discerned from the map alone that, for example, the rebels path up the river after their attack on Fort Haldane indicated that their objective was “control of the commercial zone along the rivers,” perhaps with the goal of “conduct[ing] trade on their own account, perhaps even keeping the plantations geared to the production of exports.”⁷ Undoubtedly beautiful and visually appealing, Slave Revolt is not immediately engaging as a narrative.

ESRI’s Story Map is a platform for a different kind of spatial narrative: not a pathway through a map, but as a series of maps, accompanied by a narrative. As a white paper published by ESRI spells out, text “plays a supporting role, with the map or series of maps taking center stage.“ Their logic is simple: “internet and mobile users aren’t willing to read very much.” SM-Journal, one of the series of templates that make up Story Map, displays text in a narrow scrolling panel alongside the map, rather than linked to points on a map, as in StoryMap JS. The text panel is only about 6 words wide, making it difficult to tell a complex story. There is some scope for interaction between text and map: links in the text can toggle layers on and off, zoom the map, and display a pop-up of a particular place. Further interactive elements have been omitted “so that users don’t have to figure out several functions.” ⁸

A fully realized example of a historical spatial narrative using this platform exists: Mapping Occupation: Force, Freedom and the Army in Reconstruction, a project led by Gregory Downs and Scott Nesbit. It documents the locations of the US army in the south from 1865 to 1880, and the reach of their presence, as “zones of occupation” (areas the army could travel to in order to address offenses) and “zones of access” (areas within which freed people could travel to bring complaints). The narrative “guid[es] the user through key stages in the spatial history of the army in Reconstruction,” with 10 sections of text, each accompanied by a map, most of which are interactive to the extent you can click on the points.⁹ The narrative incorporates some images, which appear as only slightly larger than a thumbnail due to the constraints of the panel. At least in my reading, the maps are essentially illustrative; while at points there are links in the narrative to toggle layers, because the links are in the narrative, they don’t encourage interaction with the map itself.

Returning to the Year of the Riot map points to the limits of these platforms. Riots are complex and multi-faceted events. While they usually have an identifiable origin, they spiral out in multiple different directions, and change shape over the course of time. Riots are not linear, so are not easily rendered in narrative or with tools that focus on sequencing points or maps. A pathway won’t tell the story in this map; nor will linear text in a sidebar, alongside a map.

Neatline, created by Scholars Lab at the University of Virginia, in collaboration with staff at RRCHNM, offers a platform with more potential for a spatial narrative of the riot than either Story Map platform. Neatline gives most of the space to a map, with a sidebar not made up of text, but of a series of “waypoints,” links that open windows on the map that can contain text or media. (Its not ideal that the popup obscures the map and can’t be dragged) The map itself allows points, lines, and polygons, as well as imported svg images. This functionality does not just allow all the forms of vector data; it effectively allows for the annotation of maps. Marking up maps offers a means of elaborating connections between points and narrative other than adding or appending text, an approach that keeps the focus on the map. Neatline also offers a basic interactive timeline, which can be used to control what is visible on the map as well as adding temporal information about what is being mapped.

In the remainder of my time, I am going offer a spatial narrative of the riot based on the map from Year of the Riot, and then prototype to what extent I can present that narrative using Neatline. I use the word prototype deliberately – working in Neatline is not part of the larger project I’m involved in. It’s a sideline inspired by the opportunity offered by this conference. So I don’t have a fully realized Neatline site to show you, only a crude mockup that skims the surface of the platform’s functionality. However, my hope is that despite those limits, using the 1935 Harlem as a case study can move forward discussion of spatial narrative.

To begin, we need to turn to what happened in Kress’ 5 & 10 Cent store on West 125^th Street in Harlem at around 2.30pm on Tuesday, March 19, 1935. Lino Rivera, a sixteen-year old Puerto Rican boy cutting through the store on his way home, stole a penknife. Two white staff members saw the theft, and grabbed Rivera. In the ensuing struggle, the boy bit both men. The commotion attracted the attention of the crowd of women shopping in the store, and a police officer was summoned.   He took the boy into an office; behind closed doors, the store manager said he did not want to press charges, so the officer released Rivera out a rear entrance, and the boy continued home. However, no one told the crowd in the store Rivera had been released, and as rumors circulated that the boy had been beaten, or even killed, the women began to protest. More police were called to clear the store, and several of the women they pushed out began screaming that a boy had been killed, drawing a large crowd, and sending rumors of police brutality spreading throughout Harlem. A group of white Communists began picketing in front of store. Twice they took to platforms to speak to the crowd, only to be dragged down and taken away by police. In the commotion someone in the crowd threw a brick, shattering one of the store windows. That violence drew crowds that spread along the long block between 7^th and 8^th Avenues, and spilled off the pavement to block the street. Mounted police, 5 radio cars and squads of patrolmen were sent to clear the street. Charges by the mounted police, and shots fired into the air, eventually cleared the crowd from 125th street, and the roadway reopened to traffic. Further clashes between crowds and the police followed, and over the next 10 hours windows were broken and stores looted across Harlem, police and passersby were injured, 4 people were killed, and over 100 were arrested.¹⁰

Initial explanations of the riot put the blame on hoodlums and Communists, before quickly giving way to the refrain that what had occurred had not been a race riot, but an economic riot: a protest against discrimination and the deteriorating conditions the Depression had brought to Harlem. The subcommittee of the Mayor’s Committee on Conditions in Harlem investigating the events of March 19 and 20 reported no evidence of any premeditation, role for the Communists, or physical conflict between whites and blacks that would warrant the label of race riot. Instead, crowds attacked property not persons. Their resentment was limited to whites who owned stores, who exploited black residents, but denied them the opportunity to work. The Committee explained the riot as the result of the strain of unemployment and discrimination, and the attitude and actions of the police, highlighting one incident in particular, the killing of 16-year-old Lloyd Hobbs by a patrolman. They covered what happened in Kress’ in detail, but devoted only a few lines to events in Harlem in the subsequent ten hours. The report contained no details of the number of deaths, injuries, arrests or damage done during the riot.¹¹

The major study of Harlem in the 1930s, Cheryl Greenberg’s Or Does It Explode, echoes that explanation, with a political inflexion. “The apparent violence against a poor black child triggered a violent outburst because it was an immediate example of two long-held community concerns: racial inequality and harsh treatment by police and white authorities.” Pointing to the focus on white property on 125^th Street, Greenberg draws a direct line to the campaigns against employment discrimination by those same stores over the preceding two years that had been ended by an injunction against picketing a few months earlier. On that basis, she argues that, “The riot was as much a political act as were organized protests and campaigns,” an outlet for passions aroused but no longer with an outlet.¹²

In the historiography of American race riots, the events in Harlem came to be recognized as the first instance of a form of racial violence that departed from the pattern established in the early decades of the twentieth century. Labeled communal riots, those riots had involved clashes between whites and blacks, generally centered in black neighborhoods, which spread over contested areas where blacks were beginning to settle, and to central business districts, where whites attacked individual blacks. When the violence directed instead at police and property within the boundaries of a black neighborhood seen in Harlem recurred in the 1940s, and then again in the 1960s, it became clear that what commentators at the time saw as something other than racial violence in fact represented a new pattern. Sociologist Morris Janowitz labeled it a commodity riot. None of the scholars of the 1935 riot offer a narrative of what occurred after the disorder spread from Kress’ store more than the few sentences offered by the Mayor’s Committee.¹³

My map of the events of the riot contains elements at odds with these accounts. It’s crucial to note that the map is not a complete picture of the riot or even of the evidence I have of specific events that took place that night. It includes at most 1/3 of the 300 or more properties damaged during the riot. I found reports of 134 men and women arrested during the riot, but the location of only 43% of those arrests. A greater proportion of the 74 individuals assaulted, killed or seriously injured, just over 75%, appear on the map. The map is also not an unmediated presentation of this evidence. To facilitate the exploration of patterns, I have organized the events into 14 categories, distinguishing acts by and against the police, assaults from injuries for which no one was directly responsible, and looted stores from those reported only with broken windows.

Taking into account those filters, more is evident on the map than attacks on white property and clashes with police.  There are at least 28 attacks by blacks on whites that fit the pattern of violence that characterized race riots prior to 1935. White storeowners, white men and women on the street, newspaper reporters and photographers, and passengers in vehicles traveling through Harlem, all suffered injuries (allegedly at the hands of blacks). Those assaults are missing from other accounts because they left few traces in the historical record: police made arrests in only 4 cases. Additional cases consequently appear only fleetingly in newspaper reports and records of ambulance callouts and hospital admissions. Attacks on whites occurred throughout the duration of the riot, providing a context of racial antagonism for the other forms of disorder.

However, that violence was more geographically contained than in race riots in the north earlier in the twentieth century: other than two storekeepers attacked in their stores in Harlem’s north, most attacks occurred around 125^th Street, with a small number further south, around the stores on 116^th Street. Moreover, there is very little evidence of the attacks in Harlem or elsewhere by whites, other than police, on blacks that characterized those riots. So while the riot involved attacks on white commercial property not previously part of public disorder involving African Americans, that is not the full story of what happened on March 19 and 20.

The map provides a basis for shaping new narratives of the riot that pay greater attention to the particularity of the events and to their spatial dimensions. It took a very specific confluence of circumstances and places needed to trigger a novel form of public disorder in Harlem. Not any report or rumor of police violence would have been enough; there were many others in the weeks and months and years preceding March 19, and the violence they provoked was focused on police. Such clashes had never extended to attacks against (white) businesses. Nor would a clash with police in the context of any business have been enough to trigger the riot. It needed to be a clash with police linked with a store on 125^th Street. As a journalist put it, for nearly a decade the 125th Street district “represented the most irritating section of white business interests in the community.”¹⁴ Only on 125^th St had stores been the target of prolonged boycotts and pickets, and continued to discriminate in employment. Only on 125^th street were there venues and businesses that had, and in some cases continued to, practice segregation as well as employment discrimination. Most of the white businesses above 125^th street had made greater accommodations to their customers, and were small, family enterprises, with few if any jobs available for white or black. Other uptown stores had moved more quickly to hire black workers in response to picketing prior to the campaigns on 125^th Street.

Only on and around 125^th St were there significant numbers of whites on the streets in addition to in businesses. It was the major shopping and entertainment district north of Central Park, and a transportation hub, catering to white neighborhoods to its west, south and east as well as the black neighborhoods to its north. At the same time, the black settlement had expanded below 125^th Street around 1930, so that the district now sat within the boundaries of a black neighborhood far larger in population and area than those that whites attacked in race riots earlier in the century. In addition, thanks to the geography of Manhattan, black Harlem was not surrounded by white neighborhoods as were other black enclaves in the city. Consequently, racial violence on 125^th Street would not draw whites from other areas of the city in the way clashes in smaller neighborhoods had. A clash on 125^th would more quickly involve white members of the CP than one elsewhere in Harlem, as they had a local HQ just north of 125^th Street on Lenox Ave, and the Young Liberators had their HQ just south of 125^th on Lenox Ave.

The map also provides a basis for a narrative of the riot that extends beyond the events at Kress, and on 125^th Street. Events spread in a complex pattern, north, further up Lenox Ave than the other avenues, and south into Harlem’s Puerto Rican neighborhoods, in several waves of activity that produced various forms of violence.While the aerial view offered by a map reveals the spread of the riot, “gazing down from a great height makes it hard to see chaos and confusion,” as Vincent Brown notes in regard to his map of the slave revolt in Jamaica.¹⁵ In the case of the riot, what we can’t see are the crowds that filled the streets for much of the night, and often literally surrounded the events that appear on the map. A point on the map generally represents a moment when groups emerged from the crowds to attack individuals, buildings or vehicles. However, crowds were present on the streets for more than the moments captured on the map.

As the riot spread over Harlem, events followed a pattern unlike any race riot that preceded it. First crowds gathered to protest, leading to clashes with police; then windows were broken; and finally, some time later, looting broke out. My map has limits as evidence of the riot’s chronology. Only 40% (69/173) of the events on the map appear on the timeline; for the majority is there no information on when they took place. That is particularly the case for stores with broken windows; only 4 of 49 appear in the timeline. Nonetheless, the map does fit with the pattern reported in other sources.

Enough time elapsed between the shifts in behavior that they need to be seen as discontinuous, and each phase of disorder considered in its own right. The progression from one to the next was not inevitable – at least, not in 1935. The period before 10pm saw a protest that expanded to clashes with police when their efforts disperse crowds escalated the violence, leading to attacks on Kress’ spreading to the other large stores that had been the targets of the boycott movement. A second phase of violence resulted after 10pm in part from police pushing crowds away from 125^th Street, on to the avenues that ran north/south through Harlem. Groups of blacks attacked whites they encountered on the street and in passing vehicles, but they encountered relatively few in those areas of Harlem. In this context, businesses provided an alternative outlet for racial antagonism. That some blacks stores were caught up in this violence is not at odds with that interpretation. As crowds moved through Harlem, not everyone would have been aware of which stores were owned by blacks. When black storeowners put up signs identifying themselves, crowds generally avoided them. Although the shift to breaking windows literally opened the way for looting, that looting did not immediately occur further suggests that attacks on stores were initially an attack on whites.

The time lag between attacks on stores and looting that began after midnight suggests that new groups joined the crowds. Given that the riot occurred in the midst of the Depression, which had hit Harlem residents particularly hard, it is unsurprising that some saw an opportunity to alleviate their economic needs by looting stores that been targets of racial violence. That the looting appears to have been concentrated on Lenox Avenue fits such an interpretation. The street was in some ways the least likely of Harlem’s avenues to be the main target of looting, as it had long been home to lower grade stores than on 7^th Avenue, but the block’s to the east were Harlem’s poorest and most overcrowded. If much of the looting was the work of the hungry, it did not supplant racial antagonism. As some groups looted stores, others continued to attack whites, in a context of intensifying violence fed by police beginning to shoot at looters.

Harlem’s riot was thus not a total break with the past. To the extent that they could, Harlem’s residents attacked whites. A spatial perspective highlights that the new forms of racial violence that appeared for the first time in Harlem resulted at least as much from the targets available to blacks as a change in their motives or the nature of racial antagonism. There were fewer whites to attack in the new, larger black neighborhood, and few whites responded to racial violence by venturing into black neighborhoods (and were less motivated to do so when provocations did not involve clashes between black and white populations, but between blacks and police and storeowners). Rather than a break with the past, the 1935 riot involved the layered violence that Dominic Capeci and Martha Wilkerson have argued characterized the later riot in Detroit in 1943: a transitional moment that “piled distinct layers of violence atop one another” to encompass both violence against individual whites, and against white authority and property.¹⁶

I’m going to turn now to my experiment telling this story using Neatline, and look at how effective my Neatline exhibit is as a narrative of the riot, and then I’m going to turn that question over to you, the audience.

The exhibit first unpacks 125^th St as a context, and then divides the events into four chronological phases, each visible when the timeline is positioned within the appropriate timespan – 2.30-10pm; 10pm to midnight; midnight to 2.30am; and 2.30am to 5.30am. (I’ve used more opaque borderless points to represent events that can’t be placed on the timeline; they appear after 10pm and remain on through the remaining phases of the riot, the timespan in which they would have occurred). That the exhibit is effectively a series of maps suggests that Neatline doesn’t handle complexity any better than the Story Map platforms. However, linking those maps and waypoints containing text and images, and linked to annotations, to Neatline’s timeline allows for a more dynamic and engaged pathway through those maps. The timeline functionality relies on an old, buggy and limited SIMILE widget that couldn’t display the number of records I had, and which I had to break to prevent from jumping forward in a way that separated records from the point to which they refer on the map. But the timeline slider does provide the means of navigating the exhibit – dragging it not only changes the points visible on the map, it also changes the waypoints visible in the right-hand menu.

Those waypoints aren’t tied to particular events; they explore the broader patterns that I’ve discussed. And rather than each being tied to an individual point on the timeline, related arguments are grouped together. The window in which each waypoint displays only holds a few sentences of text, so grouping them allows scope for more complex ideas to be developed. By the same token, breaking those ideas into smaller sections makes them more accessible, with each section title operating as a marker and a summary (when compared to a larger block of scrolling text). Groupings of waypoints also provide some flexibility in how a narrative is read – you can roll over each waypoint in a group, and explore them out of sequence, with those options limited to reduce the change that you could lose your way in the argument. And of course, opening the window on the map keeps attention on the map, and brings an element of the idea of a pathway through the map (although the tendency of the window to obscure the map negates some of that functionality – there is scope to style a theme that adjusts that positioning).

Beyond the text, the waypoints work to provide a pathway through the map in two other ways. Each waypoint can be associated with a zoom level centered on a specific location. Clicking on a series of waypoints can thus move you around a map. Annotations can also be attached to each waypoint. I’ve tried to use polygons and lines to direct attention to the analysis of space in my narrative: to movement, direction, proximity, connection, and patterns. Used in that way, annotations shift some of argument into a visual form. (I also wanted to be able to add text annotations, and to have images appear on map rather than just in record window) Combined, zooming and centering and annotations offer guidance on how to read the map, while allowing the map to retains some complexity by keeping all the points from a timespan visible – rather than having to disaggregate event types into layers, as I had to earlier to highlight patterns in the map from Year of the Riot. Having the visibility of the annotations tied to the timeline also ensures the map is not cluttered and made unimtelligible. However, Neatline does not make it possible to integrate points and waypoints – only one record can be associated with a waypoint. So while can associate many points and polygons with a waypoint, each cannot have its own record – so I could not link multiple events to a waypoint without losing information about each. (Waypoints are not an overlay on other records, but a record on par with them). As a result, the narrative that unfolds in the waypoints is essentially disconnected from the information about individual events on the map.

More functionality for Neatline is on its way. In August, the National Endowment for the Humanities Office of Digital Humanities awarded Scholars Lab a Digital Humanities Implementation grant to update and enhance the platform. This is important because it is an opportunity for humanities scholars to build for humanities scholars. I don’t think its any coincidence that I found Neatline more useful than ESRI’s Story Map, which is the product of a different community and set of disciplinary and market orientations. I’m particularly pleased that the timeline is at the top of Scholars Lab’s list of projects – both because it had me tearing my hair out as I worked on my exhibit, but also because I think its key to allowing a spatial narrative platform to offer flexible sequencing. Ultimately, the spatial narrative platform that I’m looking for goes even further than Neatline in bringing the narrative on to the map – in from the sidebar, off the slide, and above the pop-up window. What do you think?