Florence Nightingale is remembered as the lady with the lamp, walking the wards at night in a military hospital. She would have found that description irritating. The work she considered important was done at a desk in London after the war, with a pen, a stack of army returns and the help of the country’s leading medical statistician, and it is the reason the British Army stopped killing its own soldiers by accident.
The instrument she used was a diagram, and she designed it specifically for people who would not read a table.
What Was Actually Killing the Army
Nightingale arrived at the barracks hospital at Scutari, across the water from Constantinople, in November 1854, with thirty eight nurses and an appointment from the Secretary at War. The building was a converted Turkish barracks holding thousands of sick and wounded men shipped across the Black Sea from the Crimea.
The conditions are famous: no soap, no basins, rats, men lying on the floor in their uniforms, food arriving inedible. What was not obvious at the time is what the men were dying of.
They were not, for the most part, dying of their wounds. They were dying of typhus, typhoid, cholera and dysentery, contracted in the hospital. In the worst month, January 1855, the death rate among patients at Scutari reached roughly 42 percent. Wounds accounted for a small fraction of it.
In March 1855 a Sanitary Commission sent from London arrived and did the unglamorous work: flushing out the sewers under the building, which were blocked and draining into the water supply, removing the carcass of a horse from the well, clearing the drains, improving ventilation. Deaths fell sharply within weeks.
The Part Most Accounts Leave Out
Nightingale did not, at the time, understand what had happened. She believed the improvement came from nursing, discipline and supplies, and she went on believing it for a while after returning to England.
It was the analysis afterwards that changed her mind. Working through the mortality returns with William Farr, the statistician at the General Register Office, she established that the fall coincided with the sanitary works and not with her own arrival four months earlier, and that mortality at Scutari had been higher than at the front line it received casualties from.
That conclusion is uncomfortable for the person who ran the hospital, and she published it anyway. Her private notes from the period are blunt about her own responsibility. Whatever else she was, she was not protecting the story that had made her famous.
Designing an Argument
The problem in 1857 was not evidence. It was an audience: a War Office, a Parliament and a Queen, none of whom were going to work through columns of monthly mortality per thousand for two armies.
The polar area diagram, which she called a coxcomb, solves that. The year is drawn as a circle divided into twelve equal wedges, one per month. Each wedge is drawn out from the centre to a distance calculated so that its area is proportional to the number of deaths that month. The wedges are then shaded by cause: a blue outer region for preventable infectious disease, red for wounds, black for everything else.
The blue dwarfs the red. That is the whole argument, and it takes about a second to see it. Two such circles side by side, one for the first year of the war and one for the second, show the blue collapsing after the sanitary commission got to work.
Two details of the construction are worth noting, because they are where most imitations go wrong. Scaling the radius rather than the area would exaggerate large months enormously, since area grows with the square of the radius, and she took care not to do that. And keeping the wedges at equal angles means time advances evenly around the circle, so the eye reads the shape as a year rather than as a ranking.
The chart is not the most efficient way to encode the numbers. A plain bar chart would let a reader compare months more precisely, and Nightingale knew it, because she published tables too. The diagram was for the ministers.
The Campaign
The chart appeared in her Notes on Matters Affecting the Health, Efficiency and Hospital Administration of the British Army, a document of about 850 pages printed privately in 1858, and in a shorter version circulated to everyone she thought could act.
She lobbied relentlessly, and she was well placed to do it. She had an audience with Queen Victoria, who described her as clear headed, and she had Sidney Herbert, the former Secretary at War, as an ally inside the government. A Royal Commission on the health of the army was appointed in 1857, and the reforms that followed covered barracks ventilation, drainage, cooking and the establishment of an Army Medical School.
The peacetime mortality of soldiers in barracks in England, which she had shown was around twice that of civilian men of the same age, fell to the civilian rate within a decade.
She was elected the first female member of the Royal Statistical Society in 1858 and made an honorary member of the American Statistical Association. She spent much of the rest of her life on Indian sanitation, army hospital design and a plan, never realised, for a professorship of applied statistics at Oxford.
The Same Decade, the Same Method
Something was happening in the 1850s to the way evidence was presented. John Snow plotted cholera deaths on a street map of Soho in 1854 and used the spatial pattern to identify a water pump. Humboldt had introduced isothermal lines a generation earlier to show climate as a pattern rather than a list of readings.
In each case the drawing is doing analytical work rather than decorating a conclusion. A table of monthly deaths by cause conveys nothing to a reader who is not already looking for the answer. The same numbers drawn as areas make the answer unmissable, and unmissable is what you need when the reader is a minister with forty other problems.
Nightingale and Snow also shared an opponent. Both were arguing against the view that epidemic disease was an inevitable feature of certain places and certain classes of people. Both used numbers to show it was a consequence of specific, fixable arrangements. Nightingale, oddly, remained a lifelong believer in miasma rather than germs, and it made no practical difference: her recommendations were about drains, air and water, and they worked whichever theory you held.
Where the Rose Went
The diagram is now a fixture of every course on data visualisation, usually reproduced in colour and rarely explained properly. Its descendants are everywhere in dashboards, generally in versions that scale the radius instead of the area and therefore lie.
What survives better is the principle underneath it. Nightingale worked out that a statistical argument aimed at a decision has to be built for the person making the decision, not for the person who compiled the data, and that choosing the form of the picture is part of the science rather than a presentational afterthought.
Seeing the People Who Did This
Nightingale sat for photographs reluctantly and disliked the ones that existed, which is a small irony for someone whose public image did so much work. The scientists of her century were, on the whole, portrayed rather than photographed, and the portraits carry a surprising amount of information about how each field wanted to be seen.
Our book Portraying Science collects more than 400 pages of those images, from oil paintings and tempera to engravings and early photographs, arranged chronologically from the sixteenth century to the nineteenth. It is a way of reading the history of science through the faces of the people who made it, including the ones, like Nightingale, whose real contribution was not the one the portrait was painted for.
No products in the cart. 