vampire pumpkins and urban densities: an essay on method in art criticism and other intellectual pastimes

Vampire Pumpkins and Urban Densities: An Essay on the Nature of Evidence


The problem with there being two Jerry Cullum blogs is that sometimes I want to write about something of use to both readerships (or equally useless to both readerships, but I want to lay it on them anyway), and I have to explain things at length that one readership already knows but the other does not.

The context of the title is a thread of discussion of a supposed Gypsy belief that pumpkins left in a pile for more than ten days turn into vampires, a belief that apparently extends to watermelons (which raised in turn a discussion of which Balkans fruit was actually being translated by these distinctly North American terms).

The beliefs were apparently all recorded by one legendary Yugoslav folklorist in the 1930s. (“Apparently,” based on a first reading of cross-references in Wikipedia.)

The natural assumption is that the first time, the Romany folk were having him on, and every time he asked another community, “Say, do you folks have any stories about pumpkins that turn into vampires?” they too were off and running. “Let’s see if he’ll go for this one. He went for it! What else can we make up?”

Now, the right way to do linguistic or folklore studies is to play extremely dumb. You never ask the question you’re trying to learn about, you ask a question that will elicit the answer you want without the informant knowing that that is the information you actually want. (This may explain why anthropologists are so often accused of engaging in espionage.)

In the case of vampire pumpkins, the folklorist ought never to have gone further than “So those are your legends of human vampires. Does anything else ever become a vampire?” And even then you can’t be sure, because when your informant is trying to make up ever more preposterous stories, there are only so many animals and vegetables and minerals in a given vicinity, and he may think of watermelons before he thinks of, say, pebbles in the road or some other comic improbability.

As the hoax perpetrated on Margaret Mead indicated, sometimes erroneous belief in the veracity of an informant can have long-lasting methodological consequences and can affect human behavior in societies far removed from, in this case, Samoa. (That debate is not quite over; there are still those who say that the young Margaret Mead was too well informed already by her mentors to have been taken in so completely regarding “coming of age in Samoa.” Others say, if I recall correctly, that it was her mentors’ beliefs that were the problem.)

It is, of course, a natural trait to make something up on the spur of the moment, even when one isn’t trying to lie. If someone holds up a hoe and asks, “What do you call this?” a quick-witted jokester may say, “In these parts we call that a galligiggle.” But a visitor who asks “What did this big house used to be?” may get the sincere answer “Oh, back in ’29 that was going to be the Gallagher mansion, but then they turned it into a hospital in the Slump before it became the veterans’ club around 1955.” Then the well-meaning informant will realize an hour later that he has conflated two different stories and the house had nothing to do with the Gallaghers, and it was the Rodericks who were building a mansion, anyway.

So many bodies of knowledge depend on a variety of investigation techniques that the word “interdisciplinary” ought either to be retired or made universal: all disciplines ought to be interdisciplinary. A researcher learning to bracket the personal and environmental variables in a scientific experiment is having to employ several different unrelated academic disciplines. Cultural theorists who cast doubt on the objectivity of science often overlook the need to explore their own personal distortions—which does not mean that the distortions they observe in science are not there. Even a bad experiment can yield valid evidence. Even a good experiment can yield valid evidence that is extrapolated to unwarranted conclusions.

Anyone who writes or, worse, edits art criticism quickly learns how universally applicable this dictum is. A constant dilemma, in a world in which words in print are limited, is how much has to be left out because to explain it would raise too many difficulties. An artist slightly misinterprets a fact or misspells a name; the result is a perfectly valid work of art that operates excellently as an autonomous aesthetic object. But the error and its interesting consequences will have to be passed over in silence because to explain it would exceed the limits of space and the reader’s patience.

The worst example of error and wrongly weighted substantiating evidence I ever experienced was the misspelling by an artist of the last name of the poet from whom she had requested the text she incorporated into her artwork. I confirmed the spelling through a wide variety of other citations, only to be telephoned by an irate newspaper editor who informed me that his intern had finally tracked down a picture of a book cover that proved that all the literary societies and book warehouses that had featured the man over the years had spelled his name wrong. I have since noticed that sometimes even title pages spell someone’s name wrong or contain a garbled subtitle; sometimes even the person’s web page, put together by someone else, will spell things wrong. (More than one of my own online biographies contain misspellings and minor errors that were not mine but that I have no power to correct.)

This is a different sort of evidence, of course, and mostly an ultimately decidable sort; I once demonstrated through the wonders of Google’s digitized copies of original texts that an anecdote reported by Bertrand Russell and universally ascribed to William James not only had originally had a different punch line than the one Russell gave to it, but had been an event reported by Oliver Wendell Holmes when William James was just a wee lad. (Actually, I made up that last phrase out of nothing and I do not vouch for the accuracy of James’ age at the time that Holmes’ book was published. “To the best of my recollection.” And I would not quote me that it was Oliver Wendell Holmes, either, until you look up the relevant thread in the joculum blog…or was it in the comments thread of the crowleycrow blog…?)

The relative reliability of oral and textual evidence is an elementary and easy part of the problem. The evaluation of perceptual evidence and the simple gathering of on-the-ground information is slightly more perplexing. David McQueen and I were discussing Saskia Sassen’s flabbergasting ability to combine many different academic disciplines in the investigation of the phenomenon of the contemporary city, and in our separate ways, each of us was agreeing that the only problem with Sassen’s data was an overreliance on unconscious models of certain cities based on the nature and quality of the data and the nature and quality of the persons collecting the data. Some of the phenomena that Sassen described as “invisible” are not invisible at all to the researchers of a certain social class who get from the airport to the city center on the cheapest available public transport and who look for churchyards in which to eat lunch alongside the workers, whose difference from those of preceding decades is evident to anyone who pays attention to styles of clothing and to the incidentals being carried by the aforementioned workers.

This sort of evidence is anecdotal and therefore not to be trusted unless supported by data. But just as though the economic collapse of 2008 was obvious to anyone who was reading the economists and supplementing the data with direct observation, the city provides clues depending on one’s position within it and how ready one is to understand what it is that one is seeing.

But it is “understanding what one is seeing” that is the problem, and a lifetime of theoretical correction is often still not enough to uncover the truly invisible parts of the problem. The computer geeks are correct in their maxim that anecdote is not the singular of data, but anecdotes are frequently all the data that we have available, and we have to correlate the anecdotes and correct for the variables, whether the issue is the effect of changing flows of capital on individual human beings or the question of whether or not anyone on earth has ever believed in the existence of toothless vampire pumpkins.