4 ms·

Textual Strategies in Infographics: Observing Minard Observing Napoléon

52 points by masswerk 8y ago

ghaff 8y ago
There's another broader question with a design of this type. Do you want a visualization that abstracts away everything except for some essential aspect of data or do you want one that embeds a great deal of information and ties it all together in a way that rewards considerable study? This does the latter. It may or may not do so accurately (hence the article) but it's certainly not a graphic that you glance at for 5 seconds and immediately grok what it's trying to convey.
- masswerk 8y ago
  OP here. – Notably, I wouldn't criticize it for not being accurate (the priority given to conveying instructive data has been underlined by Minard himself and may be considered part of the methodology). The interesting side to it is that it is opinionated (which is probably a property of at least some types of infographics in general, since selection necessary introduces bias), the even more interesting aspect is that the opinion (or narrative) conveyed by the entire sheet is quite the opposite to what it seems to be for the chart in isolation. And, I'd suggest, it has been particularly crafted for this. You're probably right regarding the time that is meant to be spent on it. I always imagined a contemporary beholder slowly following the path with her/his finger, considering and commemorating the corresponding narrative. While it is comprehensive, it's also slow.
  - ghaff 8y ago
    Fair enough. Opinionated is a better word than accurate here. But they're often conflated. Or it just tells a specific story--which I'd argue is the goal of pretty much any infographic.
    - masswerk 8y ago
      Which separates it from scientific visualization, which is expected to be a representation as neutral as possible... (The claim of infographics is particularly a different one.)
  - pvg 8y ago
    I always imagined a contemporary beholder slowly following the path with her/his finger, considering and commemorating the corresponding narrative. You also mention something in the article about the events being in recent enough memory to not require dates. I can't imagine a non-history-nerd person looking at a similarly under-annotated map of, say, Rommel's North Africa campaign and recalling much detail of specific dates or precisely what took place at Tobruk or the Kasserine Pass or El Alamein. Would 19th century viewers have been that much different?
    - masswerk 8y ago
      It probably depends. I guess, Stalingrad is still in common memory, very much for the same reasons. Edit: This was really a major, paradigmatic event. E.g. pretty much over all of the 19 th century it was clear and went without saying, if there’s a soldier in a painting moving from right to left, it’s Napoleon’s army on retreat from Moscow.
      - pvg 8y ago
        Oh, no doubt. Hell, the nearest big street where I grew up was called 'Battle of Borodino'.
mikehollinger 8y ago
I have a copy of Tufte's "Beautiful Evidence" at work. It's a survey of different methods of conveying information, with pros and cons. I occasionally just page through it while contemplating some UI widget or another, and actually wrote a rather nice CLI tool for displaying some sensor data (histograms, auto-ranged, FTW!) based on some of his ideas. The book not only goes back through various historical treatises like the rise and fall of Napoleon, but also discusses newer ideas like "sparklines," and even addresses why bullet points and font sizes contributed to the space shuttle Columbia disaster. It's definitely worth checking out Tufte's site [1], if you're into this sort of thing. He's even got a series of small treatments on completely random topics in the "Notebooks" section. [1] https://www.edwardtufte.com/ https://www.edwardtufte.com/

52 points by masswerk 8y ago

There's another broader question with a design of this type. Do you want a visualization that abstracts away everything except for some essential aspect of data or do you want one that embeds a great deal of information and ties it all together in a way that rewards considerable study? This does the latter. It may or may not do so accurately (hence the article) but it's certainly not a graphic that you glance at for 5 seconds and immediately grok what it's trying to convey.

masswerk 8y ago
OP here. – Notably, I wouldn't criticize it for not being accurate (the priority given to conveying instructive data has been underlined by Minard himself and may be considered part of the methodology). The interesting side to it is that it is opinionated (which is probably a property of at least some types of infographics in general, since selection necessary introduces bias), the even more interesting aspect is that the opinion (or narrative) conveyed by the entire sheet is quite the opposite to what it seems to be for the chart in isolation. And, I'd suggest, it has been particularly crafted for this. You're probably right regarding the time that is meant to be spent on it. I always imagined a contemporary beholder slowly following the path with her/his finger, considering and commemorating the corresponding narrative. While it is comprehensive, it's also slow.
- ghaff 8y ago
  Fair enough. Opinionated is a better word than accurate here. But they're often conflated. Or it just tells a specific story--which I'd argue is the goal of pretty much any infographic.
  - masswerk 8y ago
    Which separates it from scientific visualization, which is expected to be a representation as neutral as possible... (The claim of infographics is particularly a different one.)
- pvg 8y ago
  I always imagined a contemporary beholder slowly following the path with her/his finger, considering and commemorating the corresponding narrative. You also mention something in the article about the events being in recent enough memory to not require dates. I can't imagine a non-history-nerd person looking at a similarly under-annotated map of, say, Rommel's North Africa campaign and recalling much detail of specific dates or precisely what took place at Tobruk or the Kasserine Pass or El Alamein. Would 19th century viewers have been that much different?
  - masswerk 8y ago
    It probably depends. I guess, Stalingrad is still in common memory, very much for the same reasons. Edit: This was really a major, paradigmatic event. E.g. pretty much over all of the 19 th century it was clear and went without saying, if there’s a soldier in a painting moving from right to left, it’s Napoleon’s army on retreat from Moscow.
    - pvg 8y ago
      Oh, no doubt. Hell, the nearest big street where I grew up was called 'Battle of Borodino'.

mikehollinger 8y ago

I have a copy of Tufte's "Beautiful Evidence" at work. It's a survey of different methods of conveying information, with pros and cons. I occasionally just page through it while contemplating some UI widget or another, and actually wrote a rather nice CLI tool for displaying some sensor data (histograms, auto-ranged, FTW!) based on some of his ideas. The book not only goes back through various historical treatises like the rise and fall of Napoleon, but also discusses newer ideas like "sparklines," and even addresses why bullet points and font sizes contributed to the space shuttle Columbia disaster. It's definitely worth checking out Tufte's site [1], if you're into this sort of thing. He's even got a series of small treatments on completely random topics in the "Notebooks" section. [1] https://www.edwardtufte.com/ https://www.edwardtufte.com/