[thesis swamp] ... find of the... day...

[kaberett]

Okay. So. For reasons! That are probably fairly obvious! While rummaging around on the internet! I stumbled across Isomorphic Relationship between Rubidium and Thallium in Igneous Minerals (Ahrens 1945), thought "oh shit why didn't I know about this paper?" and clicked through to the pdf.

1. Turns out I didn't know about this paper because it's actually about one column, and is essentially the conference-abstract version of Ahrens (1948).
   
2. But. As someone who compulsively reads any text in front of them. My eye was inexorably drawn. To the material that filled the rest of the column. Entitled Distribution of Wars in Time (Richardson 1945). I reproduce the text beneath the cut, for your convenience.

Distribution of Wars in Time

A statistical regularity in the dates of wars has been brought to notice by the following numerical process. A list was prepared of wars in the world as a whole. Each calendar year was thereby characterized by the number, x = 0, 1, 2, 3, 4, . . . , of wars which began in it. Next, the number, y, of years which had each such character was counted. A similar procedure was applied to the beginnings of peace. Here are some results:

	x	=	0	1	2	3	4	>4
War,	y	=	63	35	9	2	1	0
Peace,	y	=	62	34	13	1	0	0
110 e-μμx/x	!	=	62.0	35.5	10.2	1.9	0.3	0.0

where μ = 63/110. The formula is the Poisson law of improbable events. Other phenomena, known to be described by the Poisson law, include the distribution in time of the alpha particles emitted from radioactive substances¹, or of deaths by kick from a horse ².

This impersonal acount of the beginnings of war and of peace contrasts with the personal details in the newspapers and history books. In somewhat the same manner the statistics of marriage contrast with a love-story in a biography. The Poisson law is statistical in the sense that it does not predict the date of any future peace or war.

The particular set of wars summarised in the above table are fatal quarrels, which caused from 10^3.5 to 10^4.5 deaths, and which ended from A.D. 1820 to 1929 inclusive³. But the Poisson law, with other constants, also describes the beginning of wars from A.D. 1500 to 1931, as set out in Prof. Quincy Wright's list ⁴.

A more critical account of these regularities has been accepted for publication by the Royal Statistical Society.

Lewis F. Richardson
Hillside, Kilmun, Argyll.
March 19.

¹ Rutherford, Chadwick and Ellis, "Radiations from Radioactive Substances" (Camb. Univ. Press, 1930), 172.
² Bortkewitsch, quoted by Pearson, K., "Tables for Statisticians", (Cambridge University Press, 1914), lxxvii.
³ Nature, 148, 598 (1941).
⁴ Wright, Q., "A Study of War" (University of Chicago Press, 1942), appendix xx.

Comments

[watersword]

I have questions but I'm not sure I can cope with the answers.

[ng_moonmoth]

You mean, like "fatal quarrels, which caused from [3162 to 31620] deaths", while leaving out the then-recently-concluded one that caused over eight million deaths? Or, for that matter, the Thirty Years' War in the broader time range, which had a similar body count. Yeah, seems like a bit of cherry-picking there. I would also wonder what the geographic range things were sampled over was, and whether incorporating the combat claiming colonialism or imperialism for justification would affect the results.

[kaberett]

Right.

I almost want to go hunt down Quincy to see what the fuck, but I suspect it wouldn't make me any happier and it would definitely be procrastination, so.

[vass]

*rubs temples*

[kaberett]

like. absolutely everything else aside. the CITATION FORMATS.

[silveradept]

That is certainly a paper.

[kaberett]

RIGHT.

[passingbuzzards]

"fatal quarrels"!!!

[kaberett]

RIGHT