As part of an ongoing project, I’ve been examining the distribution of named places in nineteenth-century American fiction. (See the link for details; briefly, my corpus contains 1000+ American novels published between 1851–75.) One of the areas I’m trying to understand is the driving forces behind literary-geographic attention. In short, why are some places written about more often than others?

I’m currently finishing an article with some results and hypotheses, part of which involves a tiny case study of two cities: New Orleans and Chicago. As it happens, New Orleans and Chicago had similar populations on the eve of the Civil War; the 1860 census puts them at 174k and 145k, respectively. (Note: Technically those figures are for Orleans parish and Cook county, but, you know, close enough.) On the first-order assumption that the presence of warm bodies (hence political clout, economic output, and a reading public) drives literary interest, we might expect the two to have received about equal literary attention in the years around the Civil War. Not so. Fig. 1 shows what we see instead.


Fig 1. Occurrences of named locations as a percentage of all US named location occurrences by year. Click to enlarge

There’s a lot of jitter in the New Orleans numbers (the red squares), but a couple of things seem clear:

  1. Through most of the period 1851–75, there’s a lot more literary attention paid to New Orleans than to Chicago.
  2. Interest in Chicago picks up meaningfully after about 1870.
  3. Interest in New Orleans wanes a bit around the same time, but only to the extent that the two cities occur at about equal rates in the last few years of the corpus.

So how do these numbers match up with changes in the cities’ populations over the period? (See Fig. 2 and note its wider date range.) In sum: Not very well. New Orleans was twice the size of Chicago in 1850, but only half Chicago’s size in 1870 (on its way to being dwarfed by the northern city in the decades to come). If population and literary attention track closely, the numbers in Fig. 1 should be much closer together before 1860 and significantly further apart after 1870. Chicago is significantly underrepresented in the corpus compared to its size.

Pop Growth

Fig. 2. Census populations of Chicago and New Orleans, 1820-1900, with fit lines. Note wider date range. Click to enlarge.

How can we explain this and why do we care? I think a couple of things are going on. For one, New Orleans is by far the older city; it was founded in 1718, compared to Chicago’s 1837. Whether or not a lot of people lived in New Orleans — and they did, comparatively speaking; it was the country’s third-largest city in 1840 and was still the tenth-largest as late as 1880 — it was an established place in a way Chicago was not at the beginning of the corpus period.

New Orleans was also an important city, politically speaking, during the period. It passed Charleston, SC as the largest city in the South by the 1830 census and (like Charleston, which is also highly represented in the corpus) was a place of major consequence during the Civil War. Chicago wasn’t a major Civil War site; it was a young (if expanding) frontier town.

What New Orleans didn’t have, though, was extreme growth. It about doubled in size between 1850 and 1880, which is pretty good but doesn’t look like much next to Chicago, which grew fifteen-fold over the same years and was by 1890 the nation’s second-largest city.

So there’s evidence in this miniature case study that population matters, but that it’s far from the only determinant of literary attention. By itself, I don’t find that fact surprising. What does surprise me a bit is the extent to which population looks like a leading rather than a trailing indicator of literary attention. It appears — I say this reservedly, given the modest extent of cases treated here — that books are written about (or are set in) places that used to be populous rather than those that will be populous in the future. My guess would be that this is part of the banality (for lack of a better word) of most fiction; books are for the most part written about places, events, and topics that have been written about before — about things and places with which readers are more or less familiar. In other words, we pay attention to the things we’re already accustomed to paying attention to. And then there’s the fact of temporal lags induced by time spent writing, editing, in production, in distribution, etc., all of which compound the apparent inertia of attention.

The reason I’m a little bit surprised by this is that I was at least partially prepared to believe that literature might be a leading indicator of demographic shifts, since it’s a lot easier to write (and to read) about a place than to pick up and move there. A strain of that fact must certainly have something to do with the enduring popularity of regional and travel literature; it’s easier to read a book than to set off on a voyage. And the association of much ordinary fiction with current events could point in the same direction, in the mode of magazine trend pieces and the like. If Chicago’s a new, hip place, mightn’t it stand to reason that people would prefer to read about it than about old New Orleans, all else being equal?

I’m sure there’s some novelty-driven interest in emerging cities and demographic changes, but at least in the case of Chicago and New Orleans, it doesn’t appear to be the dominant factor driving literary attention. Much more on this problem and on related issues to come in the near future.