The only time I feel comfortable in large groups of people is when most of them are dead. A couple weeks ago, as I milled through Calvary Cemetery, noting births and deaths, epitaphs and adornments, I reflected on how lucky I was to be doing my favorite kind of activity in my favorite kind of place, and to be getting academic credit for it. Reading the various grave-markers, I made up stories about the lives of all the people interred beneath me; their relationships to one another, how far away their place of birth was from their place of death, and particularly, how did they die? I don’t think anyone can stroll between hundreds of gravestones without wondering how their owners died. For this reason, I focused on death frequencies as I analysed the data collected by myself and my peers at Calvary.
The first thing I looked at was the general death frequencies by year. There are two distinct peaks: one around 1940, and the other within the last five years. Now, I am a bit skeptical as to how meaningful any of the data examined and presented here is meaningful, due to the extremely small sample size. However, another reason I chose to focus on death frequency, was that, although there may have been a selection bias that effected which stones were recorded, the actual dates themselves are reasonably objective. By this I mean that the criteria for what is recorded as “1945” by me is the same criteria that made a colleague record “1945”: the gravestone says the individual whose death it marks died in 1945. Because of this, I feel comfortable making a prediction about the cause of these two spikes in the data, and this prediction is (you guessed it) war.
1940, as we all know, is the beginning of US involvement in World War II. Thousands of soldiers were shipped off, and thousands of soldiers died. I would not be remotely surprised if the cause of the Calvary spike around this period is somehow related to this war.
Now my suggestion for the most recent spike is also related to the war, though perhaps a little further removed: the Baby Boomers are dying. The average lifespan of an adult male is somewhere around 70 years old. After the war, when all those spry sweethearts hopped into the post-war bed they popped out a bunch of babes. Well, these babes are now reaching their seventies, and, statistically, it’s about their time to pop back off.
As I’ve mentioned before, this study came packed with bias and oozing inadequate sample size. I am very curious to see if, upon more detailed and further analysis, these same frequencies would hold true. If they don’t then perhaps they reflect that the areas of the cemetery that we examined were simply utilized the most during the peak time periods because of spatial or organizational issues.
Just for fun, I also looked at the months in which people in the cemetery died and, surprisingly, found that most deaths in our data of both men and women occurred between July-August. My only hypothesis for this is extremely unscientific: perhaps, after the cold Seattle winter, perhaps they just couldn’t take the heat.
Excellent work, I especially love the data visualizations 😉 Did you consider seriation analysis of the gravestone designs? There’s a package for that of course, several in fact, I’ve listed some here https://github.com/benmarwick/ctv-archaeology
Excellent resources, Ben. Thanks for sharing! Ian had shared the how-to guide for creating a seriation in R.
Thanks Ben! I actually did do a seriation for the gravestones for the larger lab report portion of the assignment, as Sara shared this code for writing frequency tables and creating seriations. If you’re interested, my full R code (including seriations and the other figures I made and did not use) is online here(be warned: this code was made for my eyes only and is therefore mildly unorganized…). The seriations I chose to make revolve around material and shape, I would be interested to do a further examination of decoration/ imagery on the headstones through time- but that would require a more homogenous dataset than the one we were working from for this lab : ).