This post is a review of The Minard System – Sandra Rendgen’s excellent book from 2018. It’s a book which showcases the full collection of Charles-Joseph Minard’s distinctive work, the most famous (and notorious) of which is displayed on the front cover. The depiction of Napoleon’s Russian campaign has been championed by Edward Tufte with the words “it may well be the best statistical graphic ever drawn” – it’s no surprise to find that this is referenced on the very first page of the book’s Preface
I’m also a huge fan of the seminal Napoleonic graphic, having used it myself as an example of great storytelling in my presentations before now. So it was a natural addition to my library to enjoy many more examples of his craft. The book benefits from being a great “coffee table” size to browse through his graphic work in detail. But it was a remark from Steve Wexler in his recent storytelling with data podcast where he referenced Minard and this very book. In discussing the importance of knowing your audience, he mentioned that he learned from this book about Minard’s background as a highly qualified engineer, and that his chart was not intended for the general public, but for “fellow engineers, deputies and local politicians.” This encouraged me to revisit this book and look more into Minard and his system (and not just at the pretty pictures I’d perused until now!)
On the graphic itself – what is little known is that it’s one of two graphics published simultaneously. The graphic concerning Napoleon’s march has gained so much exposure that it’s not widely known that there is a sister chart comparing it with Hannibal’s march across the Alps in 218 BC. This is shown above the title in the photo above. I certainly had no idea about this until I read the book (it hadn’t even registered from the front cover until I re-examined it after learning about the additional chart) so it was clear from the offset that there would be plenty to discover.
The book is similar in structure to Manuel Lima’s books on Circles and Trees already reviewed, inasmuch as the catalogue of visualisations is included in full to browse through with annotations, details and explanations, but this takes up the latter three-quarters or so of the book. Every chart has a detailed explanation adding background and context, which makes for a rich read even when browsing through the graphics, as I’m sure many will do. However, the introduction section is thoroughly recommended to get the most from the book, to understand the context of Minard’s work, his system, and the nature of the following graphics.
We learn how Minard developed his flow maps over his career in engineering, starting with what he termed “graphic tableaux” or “figurative tableaux”. At this point he had an early appreciation for data visualisation, stating that “the numbers – of an indisputable statistical utility – are not as easily apprehended by the eye as figures proportional to them”. Minard produced many such “flow maps” – unlike the Napoleonic graphic, the majority focused around more prosaic subjects such as flows of trade or traffic. And referring to his maps, he states that “The aim is … to make relationships quickly apparent to the eye, relationships that are instantly grasped where numbers would require the mediation of a mental calculation … in the spirit of the century in which one seeks to save time in all ways possible”
But one thing that really struck me was how he took this one step further. Many of his maps were accused (perhaps euphemistically) of being “non-Euclidian”. He would almost always use the term “approximation” somewhere in his descriptions, and would knowingly sacrifice geographical accuracy in order to show analytical accuracy in a graphical manner. We need look no further than the back cover for this.
This is part of a graphic (which in full extends further west to a rudimentarily-drawn USA) depicting European cotton imports in 1858 (and published in 1861). But in order to show the extent and volume of trade routes, he has vastly widened the straits of Gibraltar and the English Channel. He has drawn country outlines in a crude manner (and his placement of Liverpool is far from accurate) and Denmark has been almost obliterated, in order for him to show trade routes to St. Petersburg. However, the flow widths are accurate to the millimetre. He sacrifices geographic accuracy in order to make what was, at the time, a pressing point about how the origin of cotton imports from the USA (in blue) was likely to change drastically in upcoming years as a result of the American Civil War. In fact, subsequent versions adding 1864 and 1865 flows have even more butchered geography, losing Scotland and Wales, and driving the continents even further apart!
The second thing that jumped out to me as I learned about Minard’s life and career was that he retired from his job as a highly decorated engineer in 1851, aged 70, the latest he could get away with working until required to retire. Despite that, the vast majority of his now published work came in retirement, so he was continually developing and refining his craft as a hobby, a passion. So for example, the graphic discussed above was created in 1861 when he was 80 years old.
So from a personal point of view, I take great satisfaction in that somebody so (rightly) revered by Tufte, a man whose principles I respect but often question and stray from (not least in this blog) gained his respect both through bending (some would say breaking) best practice rules in order to tell a story through data, and continually honed his personal data visualisation “brand” through passion projects and working outside of his day job. I always knew I loved *that* Minard chart, but it wasn’t until I read the Minard System that I gained that extra respect for Minard, his thought processes and his approach to visualisation. Whatever you’re looking for, or expecting to find in this book, I’d recommend it, for the visualisations, and for much more beyond.