Linear History: Experimenting with Timeline Tools
Timelines are a popular way to present historical events and periods – and not a recent one. In 1769, the British scholar Joseph Priestley published A New Chart of History, building upon A Chart of Universal History by the cartographer Thomas Jefferys in 1753.1For more information about the history of visual representation and infographics, see: Nowadays, with the “digital turn”, new opportunities arise. Many virtual exhibitions, for instance, include timelines. From a historian’s perspective, they are a practical way to provide a general overview of historical key moments to the non-academic public; however, they usually serve more as an orientation than to gain new insights, and impose certain limits.
Thinking of time as a linear concept is only one approach among others, i.e. a Western one. Other cultures do not necessarily incorporate a linear thinking. In one of my more recent readings, the historian Laura Spinney mentioned Terence Ranger’s observations on telling history in her book about the Spanish Flu in 1918. A linear or vertical narrative of such a condensed event would not suffice: “what’s needed is something closer to the way that women in southern Africa discuss an important event in the life of their community,” Spinney wrote in the introduction. Ranger, himself referring to the Zimbabwean novelist Yvonne Very, explained: “[African grandmothers, mothers and daughters] describe [a particular event] and then circle around it, constantly returning to it, widening it out and bringing into it past memories and future anticipations.” The German historian Achim Landwehr also discussed the question of time and its representation: “The academically established historiography and with it the entire historical thinking that has started in the so-called Western world, is, in other terms, still biased by an understanding of time that originates in Newton’s ideal of an absolute time.”
As I grew up in a Western European society, (horizontal) timelines appear to me as the most obvious approach. The purpose of the present blog post, however, is not an analysis of how various cultures approach storytelling: it is not a theoretical essay about how we could imagine time differently. In this blog post, I relate my practical experiences with timeline tools. Until many months ago, I had not given serious thoughts to the creation of timelines. In my thesis about the history of cultural policy in Luxembourg from the 1920s to the 1970s, I mention many events and dates, which could be extracted and potentially visualised on a timeline. Thus, I did not create timelines with the aim to enhance my research as such. They were based on already existing information.
As I am not a programmer, I needed a pre-existing tool to create timelines. I was confronted with a rather broad choice of (online) tools. While experimenting with some of them, I was struggling with a series of practical questions and choices. In my thesis, I wanted to include a simple timeline for each period (interwar, German occupation, post-war) with a clean visualization. I had several, initial expectations:
- Define markers as events or periods;
- Possibility to visualise events or periods with approximate dates;
- Create different layers and categorise markers (thus improving the “reading” of a timeline);
- Colour-code markers and layers;
- Convert the timeline to a PDF or a picture file.
The first step consisted in identifying all the important events in my doctoral dissertation, in addition to the administrative evolution (ministries or ministers responsible for culture). Once I had a list, I attempted to categorize the information. In the timelines in my thesis, I defined layers labelled ministers, legal framework, initiatives and events, and diplomacy. The choice of these categories was motivated by two main reasons: I would need enough events for a category to legitimise its inclusion; and the categories should be defined in such a way that overlaps could be avoided as much as possible. I was thinking about creating a category “cultural institutions”, for instance, but this category could encompass both laws and events. The fact that my case study, the State Museums, would not be prominently displayed in these timelines partly motivated me to look for an alternative, focusing only on the institution – to which I will get back further below.
For my general timeline, I tested several tools and started with PowerPoint. Though I experimented with a plugin, Office Timeline, the result was not what I had hoped for, at least from a historical perspective and confronted with the expectations I had. The obvious lack of a dynamic zoom function, for instance, caused limitations as to the number of markers to add to the timeline. The simultaneous presence of events both with exact and approximate dates was also problematic. The timeline encompassing all three periods was quite messy (see illustration below), even though I had already made a first pre-selection of the information to include. Even as I created a timeline for each period in my thesis, I was still not quite satisfied. Pictures could not be included, and the labels of markers could not be too long. One event was completely out of frame (in PowerPoint, it appeared outside of the slide). I was not able to categorise events, and my choice of layers was not necessarily useful – partly my own fault, partly due to how the tool works. Of course, the plugin offers various templates to choose from, but the one I used was, in my opinion, the most suitable to a historical timeline. The other templates were more adapted to create roadmaps for businesses, for instance.
I particularly value the possibility to include sources (pictures or videos) and descriptions and hyperlinks, features which could only be found in an online tool. Virtual or dynamic timelines offer options that “analogue” ones are missing. Though I was aware that my thesis manuscript could only include static timelines, I hoped that the use of a virtual timeline and its exportation as a picture file might be more promising. Meanwhile, I was thinking about what expectations a virtual timeline should fulfil, in addition to the ones listed above:
- Insert or embed audiovisual material (sources);
- Add descriptions, hyperlinks, source credits (popping up when clicking on a marker, for instance);
- Dynamic zoom feature;
- Embed the timeline in a webpage (such as WordPress).
An exchange with my colleague Lars Wieneke made me aware that there was something else that I had entirely missed, particularly from an institutional and long-term perspective: the possibility to download the data as a CSV-file, in case the tool would be discontinued or the timeline itself would be deleted. Many online tools provide such an option, but the amount of data included in the CSV can vary. A CSV file would need to include all the markers, the dates, the descriptions, the layers/categories and, in the case of embedded sources, the hyperlinks.
I imagined a perfect tool for my purposes. The reality was quite different. It was impossible to find a tool that could satisfy all my expectations as a historian. Furthermore, I realised that there are, broadly, two types of timelines, though the boundaries are not always neat. There are story-driven timelines, on the one hand, and purely factual timelines that serve as a collection of key moments in history (such as the one I created in PowerPoint), on the other hand.
For story-driven timelines, Knight Lab’s TimelineJS is probably the best tool I have encountered so far, and it is entirely free, unlike most other tools that usually offer a limited free version. In simple terms, TimelineJS is a crossing between a slide presentation (a slide for each marker, including the possibility to embed an illustration or video) and a timeline (used to navigate between the slides). The timeline can be divided into different layers (defined by categories to which events are attributed). Instead of working in a user interface where the input of information is instantaneously translated to the timeline, the user works in a Google spreadsheet (saved in a Google Cloud), for which a template is provided. It needs a bit of tinkering to get used to it, but the advantage resides in the fact that this downloadable spreadsheet is also the backup and includes the entire data used for the timeline. Audiovisual material uploaded to Google Drive can be embedded in TimelineJS by including links in the corresponding columns in the spreadsheet, provided that the file formats are supported. It is also possible to embed videos from the most common video platforms (YouTube, Vimeo, etc.). Unfortunately, the timeline cannot be converted to a PDF or picture format. Furthermore, the long-term management of the content on Google Drive needs to be considered when using the tool in an institutional environment and for research.
The best premise to use TimelineJS is a coherent and chronological narrative with sufficient audiovisual material. I considered it as a fitting tool to tell the history of the State Museums. The result can be seen below. TimelineJS works well with a mix of precise and approximate dates. As the focus really lies on the slides and not on the timeline as such, this barely weighs in the user experience.
In my case, the audiovisual material is stored on Google Drive, and I needed to make sure that pictures were not copyright protected, especially the portraits of individuals. It is possible to add the source credit and a description. Unfortunately, the tool lacks an integrated PDF reader. For documents of several pages, such as letters or legal texts, I needed to circumvent this limitation by converting each page of a PDF to an image file, then collating all images into a single image file.
My quest did not end there, however. I still needed a tool that could produce clean timelines as illustrations for my thesis. I turned to the second type of timelines: a simple “collection” of events. Depending on the tools, the user can also add descriptions and embedded material for each marker. This type of timeline is far more common. When it comes to virtual timelines, though, the difference between story-driven and factual ones is not always very neat. For historians, it is difficult to find the most suitable tool; it also depends on how much money one is willing to invest. Among the more interesting ones range Tiki-Toki, Timetoast and Preceden, all of which offer some limited features in their free version. Timetoast is rather a mix of the two types described above, as it includes two visualisation options: a timeline and a list, the latter enabling a story-driven approach and automatically showing descriptions or texts for each marker. The list-mode has the advantage that events with approximate dates (only the year) are less disturbing than on a traditional, horizontal timeline. Tiki-Toki is less user-friendly than other tools I have encountered, but it offers more options. Both videos and pictures can be embedded, for instance. Personally, though, I think that the presentation is less clean. A possibility to export the timeline in JSON, PDF and CSV formats is provided. An image file of the entire timeline and its markers is not available. The zoom-function is not continuous and the zoom level needs to be chosen within a menu. Markers can be categorised and colour-coded. It is possible to define markers as periods, but they cannot be distinguished from events at a first glance. Tiki-Toki includes a 3D visualisation of the timeline, which reminds me a bit of the lanes in the videogame Guitar Hero.
Neither Tiki-Toki nor Timetoast should be entirely dismissed. It depends on the expectations of the user. I would say that Timetoast is better suited to pedagogical purposes in classrooms, for instance, as it is more accessible and easier-to-use. In Tiki-Toki, I found myself fumbling through the options. Eventually, I chose Preceden, though I had initially dismissed it for reasons that I cannot recall in hindsight. For my purposes, I subscribed to the paid version to remove the limitation of the number events on a timeline. For all the other online tools mentioned in this blog post, I used the free version. Preceden does not include as many features as other tools do, but when it comes to the core aspects of a timeline – the markers and layers – it works smoothly. Markers and layers (categories) can be colour-coded, events with approximate dates are supported, pictures to illustrate markers can be included. It is possible to connect markers, for instance when an event is linked to another one. The tool automatically rearranges the markers so that overlaps are avoided, but the user can also do it manually. Markers can be quickly modified in the bulk editor; a legend can be added to explain the colour-coding (it might even serve as an alternative to defining layers/categories). In the bulk editor, the user can disable markers to hide them from the timeline. I used this feature to create a version for each period in my thesis based on the same online timeline. A zoom function is missing, however, or at least I was not able to find one. The width of the timeline is limited to the width of the browser window. The larger the screen, the better the user experience.
I defined four layers: ministers, legal framework, initiatives and events, diplomacy. I could also have chosen others, but with this division I avoided issues with markers potentially fitting into several layers. The limited number of layers also helps a lot, and I would recommend not defining more than five layers. For the content, the post-war period was the biggest challenge because of the number of events that could be potentially shown on the timeline. The resulting timeline is quite stretched when converted to a PNG file, but it still offers a clean presentation. I could include nearly all the key moments, instead of making a choice based less on epistemological reasons than on technical ones.
Timelines can be useful additions to illustrate historical research. They might provide a general overview, or, as in the case of story-driven timelines, present a narrative in a different way. However, they can never provide a stage for the complexities historians should strive for in their research. Timelines might simplify, boil down, condense history, but they do not suffice to mediate a comprehensive view of historical reality. In 1936, the founding director of the Museum of Modern Art in New York, Alfred J. Barr, created a graphic on the history of cubism and abstract art. But it was a distinctively Western perspective. In 2019, the artist Hank Willis Thomas updated Barr’s chart and titled his adaptation Colonialism and Abstract Art.
Learning about the limits of timelines is as much important as knowing their context of creation and the choices that were made. These issues need to be considered and clearly voiced. Our historical understanding should not be dependent on the tools we have or the limits they impose. Timelines should not create the impression of teleology, of historical necessities, of inevitability. Ideally, research should not be adapted to the tools, but the tools should be adapted to the needs of historians (and users in general).
Add. 1: While wrapping up the present blog post, I discovered by chance the tool ChronoZoom, free to use and a collaboration between Microsoft Research, UC Berkeley and Moscow State University. It was designed to explore “big history” – not only several centuries, but even thousands and millions of years, from human history to a geological and cosmological scale. Unfortunately, I was not able to create an account and could not test it. It does not seem to have been maintained over the last years.