Thread by @ProfPaulPoast

1/ Today is class 2 of "Quantitative Security"!

Today, we're starting our "Intellectual History" by covering Wright and Richardson! Here's a summary

2/ We actually start with sociologist Pitirim Sorokin

en.wikipedia.org/wiki/Pitirim_S…

3/ Sorokin published a study quantifying war in 1937.

books.google.com/books?id=3A7TA…

4/ Why did he undergo (in Vol III) a quantitative study of war? On page 264, he writes...

5/ "In the immense literature on the evolution of war...exceedingly few authors have tried to consult the sources of historical facts. Of these authors...only one or two attempt to study the problems systematically and quantitatively, for several countries and several centuries"

6/ Sorokin then has a footnote where he identifies Quincy Wright as that "one" scholar.

But at the time of Sorokin's writing, Wright had only given a series of lectures on the topic at @UChicago in 1933 (published in a small book in 1935):

books.google.com/books/about/Th…

7/ It wasn't until 1942 that Wright published his magnum opus "A Study of War":

press.uchicago.edu/ucp/books/book…

8/ Note: the link is to the "Abridged" version. The original version was just massive.

Here it is -- all 1,700 pages -- on my desk:

9/ Why did Wright start such an undertaking? He explains it in Appendix I:

"These volumes bring to an end a study of war begun at the University of Chicago in 1926. In the spring of that year, on the initiative of Professor Charles E. Merriam, several members of..."

10/ "...the departments of political science, economics, history, sociology, anthropology, geography, psychology, and philosophy met together and discussed topics for research on the causes of war"

@UChicagoSSD made grants. These started the "Chicago Project on the Study of War"

11/ Side note: Wright writes, "[The money allowed] Twenty-five research assistants, mostly of graduate student level, were employed for periods of one or more years"

Those students were the first class of @UChicagoCIR!

12/ Why did Merriam call for this meeting of UChicago faculty in 1926?

In the Forward to "Study of War", Wright writes, "This investigation, begun in the hopeful atmosphere of Locarno.." which is in reference to the 1925 Locarno Accords

en.wikipedia.org/wiki/Locarno_T…

13/ At the time, many viewed Locarno as the true "End of the Great War". People were hopeful and wanted to know if data supported their hope.

14/ Given the book's length, the book (especially the appendixes) contains an overwhelming amount of data (which scholars have subsequently drawn upon).

👇is part of perhaps the book's most widely used table -- Table 31 in Appendix XX, The List of Wars:

15/ Lewis Fry Richardson was one of the first people to make use of these data

en.wikipedia.org/wiki/Lewis_Fry…

16/ Notably, he published the following JASA (@AmstatNews) article drawing on these data

tandfonline.com/doi/abs/10.108…

17/ In that article, he showed that data on all types of violence follow a "Power Law" distribution

18/ This finding has gone on to influence a host of research on the study of conflict and violence. For example, this @nature (cc @JennRichler) piece by @sgourley @Michael_Spagat & others:

nature.com/articles/natur…

19/ Richardson flies through a bunch of math and assumptions to end up with the following formula:

20/ But even he acknowledges that this is only a "suggestive formula". Did you notice the dotted line in his power law graph? That's called "EXTRAPOLATION"

Reminds me a bit of drawing an owl:

21/ After covering this history, we end class by using war fatality data from the Correlates of War project to see if those data follow a power law

correlatesofwar.org/data-sets/COW-…

22/ Turns out, most of what you need to do can be done in Excel...no need (at this point, at least) to enter into the whole @Stata v @rstatstweet debate!

23/ So that's what we'll cover this week in "Quantitative Security".

Next week, we'll tackle more directly a question underlying Wright and Richardson's work -- is war declining in frequency and magnitude?

(end)