After that somewhat depressing post about Wasilla, I’m delighted to be presenting some maps of Amherst, Massachusetts. Before I do, a few methodological and philosophical points are in order.
First, we don’t take the ascriptions in obituaries at face value. We realize that people aren’t described 100% accurately in these texts. An obituary, like many other texts, tells you at least as much about its author as its subject. We’re therefore treating these documents as reflections of what the people in a community value. Whether the deceased actually embodied all of the traits ascribed to them is not for us to say. Regardless of the answer to that question, the constellations of qualities ascribed in obituaries tell us what the friends and family of the deceased think is good and important enough to bother attributing.
There are other caveats to consider. For instance, the vast majority of the people celebrated in obituaries are adults in their 60s and above. So these texts tell us about what various communities value in the elderly. Whether they also value such attributes in the young and middle-aged is an open question.
Additionally, as Dana Rognlie, a terrific graduate student in philosophy here at UO pointed out to me recently, we shouldn’t presume that the authors of obits are a random sample of the local community. Presumably, they’re almost all close family or friends. But which family and friends are they? Are they usually the daughter, the son, or the spouse of the deceased? Or are they typically collaborations among all of the close family? If it turned out that 80% of obituaries that were written by a child were written by a daughter, that would be good to know. Unfortunately, we don’t yet have any data on this, but we’re looking into it.
Next, we don’t think of these maps as comprehensive. In particular, we think that a trait is considered a virtue in a community if, but not only if, it tends to be attributed in the obituaries composed by members of that community.
We also think, with Hume, that distinctions among intellectual, moral, political, and other kinds of virtues are blurry at best. The rich array of thick terms used to describe the dead doesn’t seem to be carved at the joints by these distinctions. One of the things most often said about the deceased is that they were a friend. Is friendship a virtue? In a forthcoming paper, I argue that it is, but I realize that that’s contentious. Another thing that’s often said about deceased men is that they were veterans. Is a group affiliation of this sort a candidate for virtue? Robert Adams thinks so, but again it’s contentious. Soldiers do things qua soldiers. Another thing that’s often said about the dead is that they were fans of the local sports team (the Ducks in Eugene, the Patriots in Amherst, and so on). Fandom is about as passive as being for the good gets. When your team wins, you’re, as Garfunkel and Oates put it, vicariously, “temporarily, adjacently victorious.”
Finally, we think that obituaries and other talk about the dead lend an interesting perspective to discussions in meta-ethics and philosophy of language. What kind of speech act are we performing when we call a dead family member generous (one of the most common terms used in obituaries)? It looks like an assertion, but as anyone who’s encountered Pericles’ funeral oration, Plato’s Menexenus, the Gettysburg address, or John Cleese’s eulogy for Graham Chapman can affirm, talk of the dead tends to go non-cognitive pretty quickly; it turns into an exhortation of sorts to the audience.
With all that out of the way, here is a map of the values of Amherst, MA:
This one is quite detailed, so I encourage you to open it in another tab and explore it by zooming and scrolling. As with some of the other maps we’ve presented, the size of terms here is determined by the number of terms that co-occurred with the term in question, not simply the number of times that term occurred. The width of an edge connecting a pair of terms represents the number of times they co-occurred. Centrality/peripheralness represents, well, centrality and peripheralness to the network. And in this case color represents modularity. Modularity is, somewhat roughly, a measure of the density of interconnections among nodes in the network. In this map, each color represents a different module; terms within the same module tend to be more connected with each other than they are with terms in other modules (represented by different colors).
The green group seems to encompass a mix of intellectual and political virtues, notably including wit, pedagogy, feminism, civil rights activism, and political engagement. This group is also the first to have a major node for a religion other than Christianity: Judaism (it also contains a small node for Islam. The blue group seems to encompass a variety of other-regarding dispositions, including humor, helpfulness, environmentalism, and compassion. The pink group seems to be primarily about commitment to the local community, including one’s family, friends, church, and civic community. The red and yellow groups are probably too small to interpret.
If you’ve been following my previous posts that detailed Eugene, Flint, and Wasilla, you’ll probably have noticed some interesting differences. This map is by far the most complex. That’s in part because I was able to look at a lot more obits for Amherst (about 600… oy). It’s also because these obits tended to be quite a bit longer and have richer descriptions. That’s unsurprising, given how much of a class and educational difference there is between Amherst and the rest of the towns I’ve looked at so far. This map also has much less focus on sports and religion and much more focus on political and intellectual engagement. Depending on your prejudices, you might find that unsurprising.
In other towns, we noticed some pretty substantial differences between the constellation of traits associated with women and the constellation associated with men. What gender differences turn up between men and women in Amherst? Here’s the map for women:
Again, these are pretty detailed, so I encourage you to open them in other tabs and explore by zooming and scrolling. No male nuns — unsurprising. No male feminists – disappointing. Fewer female sports fans — unsurprising. No female veterans — unsurprising. Otherwise, there aren’t that many noticeable differences between these maps.
I’ll post a “complete” map comparing attributions to men and women in all towns surveyed so far in a later post. For now, I need to take a bit of a break from reading obituaries….