The history of statistical graphics and thematic cartography are intertwined with each other, and with the rise of statistical thinking throughout the 19th century. Over this time, social, moral, medical, and economic statistics began to be gathered in large and periodic series, and the usefulness of these bodies of data for planning, for governmental response, and as a subject worth of study in its own right, began to be recognized.The birth of statistical thinking was also accompanied by a rise in visual thinking: a desire to make these numbers-- their trends, tendencies, and distributions--- more easily communicated, or accessible to visual inspection. This session brings together a collection of topics representing notable landmarks in historical developments in statistics and graphics.
The presenters are members of ``Les Chevaliers des Albums de Statistique Graphique'', referring to the Albums of the Ministry of Public Works of France, 1880--1899. Each presentation will focus on one aspect of the genesis of important ideas in statistics, graphics or thematic cartography.
| Speaker | Title |
|---|---|
| Ruediger Ostermann | August Friedrich Wilhelm Crome: A german pioneer of graphical statistics |
| Ian Spence | Le camembert depuis 200 ans: Playfair, Minard, et les autres |
| Michael Friendly | Visions and Re-Visions of Charles Joseph Minard |
| Antony Unwin | Galton, Statistics and Graphics |
| Gilles Palsky | Paris in numbers: Jacques Bertillon and the first parisian statistical atlas |
| Forrest Young | Statistical Graphics: What's New and What's Not! |
| Antoine de Falguerolles | The "atlas graphique et statistique de la Suisse" of 1897 |
| Theodor Porter | Graphical Representation in the History of Statistics |
Charles Joseph Minard is most widely known for a single work, his poignant flow-map depiction of the fate of Napoleon's Grand Army in the disasterous 1812 Russian campaign. In fact, Minard was a true pioneer in thematic cartography and in statistical graphics, and developed many novel graphics forms to depict data, always with the goal to let the data ``speak to the eyes.'' This paper reviews Minard's contributions to statistical graphics, the time course of his work, and some background behind the famous March on Moscow graphic.We also look at some modern re-visions of this graph from an information visualization perspecitive, and examine some lessons this graphic provides as a test case for the power and expressiveness of computer systems or languages for graphic information display and visualization.
The last quarter of the nineteenth century saw a growth in the number of statistical figures, maps and diagrams in official publications. They answered the wishes expressed at the international statistical congresses (The Hague, 1869, or St-Petersburg, 1872). In France, the Album de statistique graphique of the Ministery of public works was published as early as 1879, under the direction of Emile Cheysson. It served as a model for other administrations.Thus the statistical service of the parisian municipality, directed by Doctor Jacques Bertillon, started upon the annual publication of an Atlas de statistique graphique. The plates constitute a unique collection of statistical representations at the scale of a big city. The publication was stopped after two volumes, because of its cost and its difficulty, in comparison with statistical tables. Nevertheless, from the diversity of its themes and graphical processes, Bertillon's Atlas finds its place among the most accomplished works of French graphical statistics.
Franics Galton was a Victorian polymath and no short talk can hope to do justice to the range of his contributions, even just to those within statistics. Nevertheless, there are some common features to much of his work which are worth emphasising and can be seen in many examples: his concern for testing theories with data, his appreciation of the importance of proper data collection, and his interest in using graphical displays. What should we think of his analyses nowadays?
A.F.W. Crome (1753 - 1833) was a very prominent man in his time, but almost unknown as a statistician today. Crome had two faces: a political one and a statistical one. He wrote several pamphlets which lead to passionate political discussions. In some of these pamphlets he used statistical methods to prove or illustrate his politicial theses.The statistician Crome publihed several interesting books. In his "Europens Produkte," he developed a statistical map which describes several European states, but in a very abstract manner.
In other statistical papers of Crome we find bar charts and area plots used to visualize data. On the latter topics he was a competitor of William Playfair. In this talk we give
- a brief curriculum vitae of A.F.W. Crome
- a short discussion of his politicial activities
- a description of his "Produktenkarte"
- an overview of Crome's graphical activities
William Playfair invented the pie chart in 1801. In the following two centuries, this serviceable and likeable little graph has found many uses and users. Charles Joseph Minard was an early and enthusiastic adopter, frequently using the form in his cartes figuratives. Since then it has become one of the most widely used of statistical graphics. However, despite its popularity, the pie chart has had to endure the disapprobation of experts in statistics and graphical design; the academy has advised avoidance of what it imagines to be an ill-conceived creation. Happily and (as we shall show) wisely, the masses have chosen to ignore the advice.
It is truly amazing to see the collection of historical graphics inspired by Minard and published by the Ministry of Public Works in France from 1879-1899. The graphics used in these publications include, in addition to flow charts like Minard's well known "March on Moscow", charts which we think of as recent inventions: star charts, pie charts, exploded pie charts, bar charts, side-by-side bar charts, time series charts 1-way mosaic diagrams, geographical charts, etc.Some particular points of interest in these works include: 1) Essentially all of the charts use color very heavily to encode information; 2) Many charts involve interesting combinations of the basic types, especially common is the use of geographic maps with other graphics located strategically to show information about geographic subunits of the map; and 3) There are many charts that are new to me (ones which might be called half-pie charts, hat-charts, etc.).
I will present numerous slides that I have made from scanning the three volumes I have easy access to, and believe the viewer will agree with my conclusion that "What's new is not much, and what's not is what's new"!
Graphical methods were in widespread use in the second half of 19th century. The 17 "Albums de Statistique Graphique" published in France between 1879 and 1899 under the supervision of Emile Cheysson (1836-1910) exemplify mass-production of high quality thematics maps and quantitative graphics. The "atlas graphique et statistique de la Suisse / Graphisch-statistischer Atlas der Schweiz" of 1897 also illustrates this current practice for presenting data.While prefaces in the former acknowledge the importance of Charles Joseph Minard's pioneering work, the preface of the latter refers the "Gutachen \"uber die Anwandung der graphishschen und geographischen Methode in Statistik", which Goerg von Mayr had published in 1874. In this presentation, I will present slides from the Swiss atlas and compare them to similar graphics from the French albums.
The history of statistics has become interesting in recent decades to a remarkably diverse group of scholars, and for an equally wide-ranging unwieldy set of reasons. Among them: (1) The incorporation of chance, and of statistical reasoning, into the sciences has involved some fundamental intellectual changes, perhaps even a kind of scientific revolution. (2) As a social technology, statistics is a form of information that is also an important medium of power. (3) Statistics joins the human sciences to the physical and natural ones. More than that, it has come to involve in a particularly striking way an ideal and an ethic of objectivity, meaning, among other things, control of the subjective. (4) Statistics has become an indispensable language of public discussion, with immense credibility even if it is also sometimes condemned.Ever since Emile Cheysson, and perhaps even earlier, graphical representation has drawn interest as a form of visualization and a means of communication, and these roles are important enough. But it's significance extends still further. In this talk I will frame and reflect on the papers in the session mainly in relation to issues of objectivity and power.