March 2nd, 2014

Worlds, not stories

Here is the video of my talk at visualized, presenting the selfiecity project:

At the end, I tried to make a more general point, which is of big importance to me, and which I would like to expand on in the following:

You will often hear these days, that data visualization is great for “telling stories”, to “make the complex simple” or to “make boring facts exciting”.

While this all true to some degree, I think it misses the greatest quality of data visualization today: to provide us with new kinds of “glasses” to see the world.


Joel de Rosnay described a fascinating, futuristic device in 1979: the macroscope. Just like the telescope allows us to see far beyond what our eyes would allow us, into the depths of space, and the microscope allows us to look at the infinitely small, a device called macroscope could allow us to investigate the infinitely complex: society, and nature.

We have this device now on our hands: data science and data visualization. It provides with new kinds of “glasses” to look at the world, a body extension that gives us the superpower to change the realities we are looking at, beyond the physical world.

This is a truly fascinating prospect, and the potential of how these new views of the world can also change our world views (cue in overview effect) is actually what keeps me motivated each and every day.

Yet, no world-spectating device is a neutral technology. Just like with photography, there is always a human behind the device, making decisions. How far do we zoom in? What is the focal point of the image? Black and white, or colored? Out of hundreds of shots, which is the one we publish? All of these decisions are made by photographers day in and day out, and the more you think about it, it becomes clearer and clearer that a photo is never an objective reflection, but always an interpretation of reality.


From the brilliant: Scott McCloud: Understanding Comics

The same is true for data visualization. In a way, I feel like a photo reporter myself. In each new assignment, I chart new territory, travel to unknown “data countries”, lift every rock, and document as much as I can.

But, when I come home, and when it is time to make decisions on what to publish, I need to think about a concise, but also scenic and sensually rich route to what I found to be the essentials of the pheomenon at hand. What were the impressions that lasted? And which images were able to capture this whole world I experienced in one shot?

And exactly in this editing process lies the main editorial contribution of the data visualization designer as author. Let’s make no mistake – even a very data-heavy, “sober” representation of data has an author who made clear decisions on what to include or not, what to combine, or not and what to prioritize. And the same holds for the underlying dataset. So, fully acknowledging the role of authorship, with all the journalisitic responsibility it brings, is an important result of this line of thinking for data visualization.

Let’s face it: the most important issues today are not photographable: financial speculation, climate change, the credit crisis, tax evasion – all these important issues cannot be photographed. Data visualization can help us both to understand these complex issues a bit better, but also to provide images to debate about, to refer back to, and sometimes just to meditate over.

This is why I see data visualization as sort of a new photojournalism — a highly editorial activity using a deceivingly objective-looking apparatus.

Now, what does that have to do with storytelling?

Well, in short, I want to turn my audience from mere consumers, who just passively listen to all the exiting stories I can tell them to fellow travellers, who learn to enjoy data exploration the way I do. I want them to use the visualizations I provide as starting points for their own explorations.

Why? Well, as I tried to sketch above, data does not always equal truth. I want everybody to become a critical consumer of information.

Second, in my view, no knowledge sticks as well as the self-created knowledge: after a first hunch, to look further into the evidence, maybe pondering counter-evidence, and finally formulating a specific, grounded belief – this is the type of knowledge you will never forget.

The way kids learn — probing a possibility space, trial and error, action and reaction — this type of active information discovery is something I really like to promote, in order for us all to become and stay critical consumers of all the information that surrounds us, and interactive visualizations can help us train that muscle.

Consequently, any serious visualization of a sufficiently complex topic should always aim at exposing the complexity, the inner contradictions, the manifold nature of the underlying phenomenon. I like to provide users with a structured way to explore a complex phenomenon on their own terms, in a sensually rich mosaic of media and facts rather than a pre-digested narrative with a surprise at the end. To me, interesting topics rarely boil down to a single story.

Of course, I am a fan of smartly applied narrative tricks to guide the audience’s attention in subtle ways. In the end, a lot of the design skills in information design boil down to establishing a certain dramaturgy that unfolds when people interact with our pieces.

But, to me, over-emphasizing the role of storytelling (demanding that every visualization should “tell a story” – whatever that means, anyways), carries the risk to throw us back into linear, author-driven, leanback media formats (think “TV”). I think there are great lean-back media products, but data visualization is best consumed wide awake, standing upright, ready to jump.

So, let’s work on exposing the complexities and ambiguities of working with data. Let us get better at knowing and questioning the origin of the data. Let us take time to investigate and explore the material, to create variations and make mistakes. Let us design for demanding, curious audiences.

Let’s tell worlds, not stories.

(Thanks Bruce)

29 Responses to 'Worlds, not stories'

Subscribe to comments with RSS

  1. Gregor
    March 2nd, 2014 at 10:39 pm

    Thanks for writing this, Moritz. I’m tempted to re-blog this post under “Why I started with data visualization”. Might actually do it…

  2. Kim
    March 3rd, 2014 at 11:39 am

    Yes me too. Very well written compendium of all the things I’m twisting my mind around at the moment. Thanks!

  3. Moritz Stefaner
    March 3rd, 2014 at 11:43 am

    Thanks guys! Glad to have finally put that train of thought into writing.

  4. Walter
    March 3rd, 2014 at 12:44 pm

    Brilliant, thanks for sharing!

  5. mprove
    March 3rd, 2014 at 1:20 pm

    great stuff! thanks.

    You’ve also managed to remind me of Scott McClouds recent session at IxD14 in Amsterdam. His lunch conversation is online at

    BTW Can you please urge the conference organizers to make the videos public?

  6. Dino Citraro
    March 3rd, 2014 at 6:00 pm

    Nicely summarized, Moritz. I’ve been having similar thoughts and discussions lately, and I’m glad you made this post.

    I haven’t noticed many clients ask for the “pre-digested narrative with a surprise at the end”, is this something you come across often? It seems indicative of a person desiring a static infographic.

    Regarding this point: “any serious visualization of a sufficiently complex topic should always aim exposing the complexity, the inner contradictions, the manifold nature of the underlying phenomenon”, I really like the intent of that, but I’m not sure that it is a realistic expectation (or perhaps even possible every time). Providing interactive access to the raw data seems to help address the need to expose the complexity, and perhaps a narrative element will help frame the inner contradictions, but given the subjective nature of the data collection process and the thinness of any datapoint (it will always lack situational context and complexity), I’m not sure the manifold nature of the underlying phenomenon can ever fully be revealed. Seems like you would need a macroscope to look at an array of macroscopes.

    Perhaps our expectation for data visualizations should simply be that they are always honest, and provide the ability to verify any conclusions or pattern assumptions?

    Congratulations on the launch of selfiecity. Fascinating stuff in there.

  7. Moritz Stefaner
    March 3rd, 2014 at 7:27 pm

    Hi Dino, thanks – great comments!

    I don’t get too many requests for these simplistic linear stories these days, but that is also related to the fact that I don’t do marketing/PR type jobs anymore. But I do still encounter situations where graphics or visualizations that are open-ended or exploratory in nature are quickly dismissed as “having no point” or “can’t see the story here” (even among professionals), and I think this is partly due to the “storytelling” overemphasis.

    About the complexity – I just think taming complexity is the natural habitat of data vis. For instance, you can make a simple, compelling, emotionally touching point with many other media, from spoken word over cartoons to illustrations, movies, etc. But no medium is as capable as data visualization in making complexity graspable.

  8. […] role that the designer plays in visualizing data (you can read them here and here). This week, Moritz Stefaner did a much more eloquent (and concise) job of underscoring the sensibility and the responsibility […]

  9. Nick Diakopoulos
    March 4th, 2014 at 3:16 pm

    I think sometimes there’s an overemphasis on narrative amongst journalists — it is the tried and true way for them to communicate information to the public. Oftentimes journalists want to simplify and clarify information for the public, acknowledging that there is a cognitive efficiency argument for them, as professionals, to do the heavy lifting of chiseling a take-away, a nugget of knowledge, for the public to take away from the data without them having to spend lots of time digging through it. So, there’s a cultural influence of the prominence of story among journalists.

    You’re of course right about the tension between story and exploration, it’s a dilemma that can’t really be “solved”, but there are ways to combine. To me it’s not about “tell worlds not stories” but rather, “depending on your audience, decide whether an exploratory or communicative visualization is more apt” – there’s no one-size-fits all here. The bests visualizations are bespoke – content, and audience, specific.

  10. Moritz Stefaner
    March 4th, 2014 at 3:40 pm

    Thanks, Nick!

    Maybe just to clarify, I am fully aware there is some really nice narrative work in data visualization, which I appreciate a lot, and which might be the perfect solution for a given problem at hand.

    I guess the main point of my posting is to make clear that data visualization does not equate story-telling, and that maybe different metaphors (macroscope, photojournalism) can help us grasp the unique features and the “natural habitat” of the medium a bit better.

    You can create great narrative also with words, video, animation, cartoons, etc. So to me, this is not the most exciting feature or a unique characteristic of the medium.

  11. Erik Jacobsen
    March 4th, 2014 at 10:21 pm

    I really appreciate this articulation of your thinking. It brings into focus some things that have been swimming around in my head since Tapestry last week.

    I think what draws many of us to data visualization is the pursuit of truth. Alberto Cairo referenced that in his Tapestry talk, though he was more comfortable using the word “truthful” than “truth”.

    There are, undoubtedly, many paths to understanding truth. I have discovered much through self-directed exploration of accessible data. I have also come to understanding through stories that resonate deeply.

    I like the idea of data visualization as a macroscope. There is so much that is invisible when looking directly at data until you can step back and see it through a thoughtfully crafted macroscope.

    Perhaps the most powerful stories function as a macroscope just as much as the “sensually rich mosaic of media and facts” that you help people navigate through your work. Think of the fables and parables that work on many levels – speaking to the reader wherever they happen to be.

    I think it’s right to be wary of stories that are too didactic, too author-driven, that tell us what to think instead of inviting us to make our own personal connection.

    I recently saw the film “The Secret Life of Walter Mitty” – it includes the (apparently fictional) motto of Life Magazine that I think applies nicely to the purposes of data visualization: “To see the world, things dangerous to come to; to see behind walls; to draw closer; to find each other; and to feel. That is the purpose of Life.”

    Thanks for advancing the dialogue.

  12. Marty
    March 5th, 2014 at 9:59 pm

    Lovely-written and compelling piece, Moritz. I don’t fundamentally disagree with any of the ideas here, but I just had to drop in to echo some of Nick and Erik’s sentiments.

    As powerful and compelling as it is to be able to truly put an entire dataset at one’s fingertips, I have a very hard time recommending exploratory vis for most projects designed for public consumption. I have seen way too many information designers (including myself!) buy into the approach of “we’ll put all the data out there and let the viewer draw their own conclusions!”… which only leads to a complicated and/or boring vis, and underinformed and/or misinformed(!) readers. It’s a cynical-sounding generalization, but I think it’s true: the public in general is busy, have short attention spans, and don’t nearly have the data literacy or domain knowledge to be able to draw insights or understanding from a hulking combination of datasets. As information designers, our jobs should be not just to transform data into a visual medium, but to contextualize it in a way the reader can understand and relate to. It’s all too easy to abnegate that latter role by shoving data into an exploratory vis. Thoughtful design can, of course, make exploratory vis more relatable and engaging, but it’s still a massively uphill battle and definitely not practical or possible for every project.

    On the other extreme, there are definitely overly didactic/shallow visualizations that over-curate data and information (cheesy infographics are always an easy target). I have a hard time blaming these on an over-emphasis on storytelling, though. Their problem is a good old-fashioned case of not being truthful: complexity, uncertainty and situated-ness is a fundamental part of any dataset. Thoughtful visualization cues the reader into that complexity, uncertainty, and situated-ness; no matter if the vis is story-driven or prosaic, explanatory or exploratory.

  13. Francis Gagnon| Voilà
    March 6th, 2014 at 3:55 pm

    Thanks for taking the time to write this down. It’s a rare post that’s truly enlightening, that enhances the way I think about data visualization. A nice follow-up to your talk at Visualized.

    I’d bring a nuance: the most important issues of our time — “financial speculation, climate change, the credit crisis, tax evasion” — can be photographed. They have been and we have some strong images: the melting ice caps, the opulence of bankers, the faces of families losing their homes. It doesn’t detract from your argument that data visualization is simply one more tool that adds a new perspective.

    Thanks for taking the community forward.

  14. Data Viz News [46] | Visual Loop
    March 8th, 2014 at 4:42 pm

    […] was also one of the hottest topics of the week, and finally, a shout out to Moritz Stefaner’s new provocative post, Robert Kosara’s new project, Simon Rogers’s article and new books about infographics […]

  15. Data Recovery Utrecht
    March 11th, 2014 at 10:51 am

    It’s beautiful how you can look at the world with bright new eyes. And sometimes, because of data shown in a different way, people just might get stuff that they otherwise wouldn’t. Which is really great!

  16. Data Viz News [48] | Visual Loop
    March 29th, 2014 at 5:34 pm

    […] data visualization keeps leading to interesting, in-depth blog posts, after Moritz Stefaner’s Worlds, not stories, that we mentioned in the previous Data Viz News. Here, Robert Kosara gives his two cents about the […]

  17. DaveZ
    April 7th, 2014 at 1:01 am

    Outstanding perspective! The way you represent the data visualization viewpoint is something I have been looking for to catalyze my own approach towards this subject. Working in an very large enterprise environment provides so much data, mostly raw, that no one quite knows how to utilize it in a sensible way. I appreciate your storytelling comment as well. My own goal in visualizing data has been to provide a singularly interpretal understanding of the data. Doing this is incredibly difficult to achieve in most circumstances. However, with some of the new tools and techniques that people like yourself are starting to push it will make our jobs easier. Good stuff!!

  18. #IC14NL: Selfiecity | Masters of Media
    April 11th, 2014 at 5:01 pm

    […] it allows the user to verify the results presented by Selfiecity. This resonates with Stefaner’s blog post claim that “data does not always equal truth” and that we should always remain critical […]

  19. […] growing field in computer science, digital media art and social science– has become an effective tool to see the world “with new kinds of glasses” (Stefaner). Again, we chase to see the world […]

  20. […] seems especially true for exploratory data visualization, Moritz Stefaner argues during his lecture ‘Worlds, not stories’ in New York, February 2014. The enthusiast scholar perceives the data journalists, academics and […]

  21. […] Stefaner on letting people explore worlds, rather than giving them pre-packaged stories: Worlds, not stories and Look, ma, no […]

  22. […] Worlds, not stories 21 comments […]

  23. top plc training institutes in india
    May 16th, 2014 at 10:56 pm

    Hey! Do you know if they make any plugins to protect against
    hackers? I’m kinda paranoid about losing everything I’ve worked hard on.
    Any suggestions?

  24. Jake Kahana
    June 30th, 2014 at 8:29 am

    […] “Worlds, Not Stories”, Moritz Stefaner, […]

  25. Data Viz News [48] - sqoops
    July 24th, 2014 at 9:52 am

    […] data visualization keeps leading to interesting, in-depth blog posts, after Moritz Stefaner’s Worlds, not stories, that we mentioned in the previous Data Viz News. Here, Robert Kosara gives his two cents about the […]

  26. […] > 7. Stefaner, Moritz. “Worlds, not stories.” Well-formed data. March 2, 2014. < […]

  27. 50 great data viz articles of 2014 | Visualoop
    January 24th, 2015 at 1:30 pm

    […] in-depth blog posts, thrpoughout the whole year, mostly after Moritz Stefaner’s post Worlds, not stories. Here, Robert Kosara gives his two cents about the discussion, and you should also read the ones by […]

  28. […] taking you for a quick tour around the world. I hear storytelling with big data is the next big thing, so we just wanted to hop on the bandwagon there :D. On a more serious note, these quick highlights […]

  29. […] Share the raw data, and turn your audience, as Moritz Stefaner says, from consumers into fellow travellers who enjoy exploring the data with […]