Thanks for a good discussion in our last class on methods for studying texts and artifacts. Lecture slides are now posted in the usual place, as are the assignment examples that I discussed in class. Note: the slides begin with a review of the some of the salient points that carry forward from our discussion of interviews, and the material on texts and artifacts begins mid-way through.
We also discussed this Nerdwriter video, "How Donald Trump Answers a Question":
I also mentioned the long tradition of structured analyses of State of the Union addresses, and although I didn't spent much time on it in class, you can read more about the topic here: http://stateoftheunion.onetwothree.net/essay.html.
But, as always, we shouldn't be too hasty to trust any given method or statistical test just because it seems easy to use and produces intelligible results. The Flesch-Kincaid grade-level test is definitely one to use carefully, or potentially not at all. Recall that the Nerdwriter video above seems to apply it fairly straightforwardly. You can find another fairly typical use of this test, accompanied by an attractive infographic, in The Guardian's story on how "The State of Our Union Is... Dumber." Yet it's also worth considering this critique of the Flesch-Kincaid test by The Economist, as well as this one (the latter was written as a direct rebuttal of the Guardian story/infographic). There's a lesson here about the power and danger of statistical methods for studying texts: the F-K tests may be tools that are as ready-to-hand and easy to use as, say, a hammer -- but as the saying goes, when all you have is a hammer, everything looks like a nail.
Incidentally, those of you who took Analytical & Historical Bibliography with me last term may have been thinking back to that course, which covered very different kinds of methods for studying texts and artifacts (and, indeed, for studying texts as artifacts). Anyway, fun fact: remember W.W. Greg? It was his gradfather who founded The Economist, mentioned just above. Greg himself was expected to take up the venerable periodical's editorship, but got distracted by the study of literature at school and went on instead to become one of the twentieth century's most influential bibliographers. (Perhaps that's only a "fun" fact by bibliographical standards...) I'd like to think he would have approved of The Economist's critiques of statistical bogusness.
We also discussed this Nerdwriter video, "How Donald Trump Answers a Question":
I also mentioned the long tradition of structured analyses of State of the Union addresses, and although I didn't spent much time on it in class, you can read more about the topic here: http://stateoftheunion.onetwothree.net/essay.html.
But, as always, we shouldn't be too hasty to trust any given method or statistical test just because it seems easy to use and produces intelligible results. The Flesch-Kincaid grade-level test is definitely one to use carefully, or potentially not at all. Recall that the Nerdwriter video above seems to apply it fairly straightforwardly. You can find another fairly typical use of this test, accompanied by an attractive infographic, in The Guardian's story on how "The State of Our Union Is... Dumber." Yet it's also worth considering this critique of the Flesch-Kincaid test by The Economist, as well as this one (the latter was written as a direct rebuttal of the Guardian story/infographic). There's a lesson here about the power and danger of statistical methods for studying texts: the F-K tests may be tools that are as ready-to-hand and easy to use as, say, a hammer -- but as the saying goes, when all you have is a hammer, everything looks like a nail.
Incidentally, those of you who took Analytical & Historical Bibliography with me last term may have been thinking back to that course, which covered very different kinds of methods for studying texts and artifacts (and, indeed, for studying texts as artifacts). Anyway, fun fact: remember W.W. Greg? It was his gradfather who founded The Economist, mentioned just above. Greg himself was expected to take up the venerable periodical's editorship, but got distracted by the study of literature at school and went on instead to become one of the twentieth century's most influential bibliographers. (Perhaps that's only a "fun" fact by bibliographical standards...) I'd like to think he would have approved of The Economist's critiques of statistical bogusness.
