By your description, it feels like the kind of book where an author picks a word and then rambles about it like an impromptu speaker. If this had an extraordinary thesis requiring extraordinary evidence like Manufacturing Consent then lots of anecdotes would make sense. But the thesis seems too vague to be extraordinary.
I get the impression of the kind of book which where a dense blogpost is stretched out to the length of a book. This is ironic for a book about subtraction.
Yup, very well-put.
Your point about anecdotes got me thinking; an “extraordinary thesis” might be conceptualized as claiming that the distribution of data significantly shifted away from some “obvious” average. If so, showing the existence of a few data points greater than, say, 4 standard deviations from the “obvious” average actually can be strong evidence in its favor. However, the same is not true for a few examples ~2 standard deviations away. Maybe Klotz’s error is using anecdotes that aren’t far enough away from what intuitively makes sense.
Probably didn’t explain that very well, so here is a Tweet from Spencer Greenberg making the point:
1. By Bayes Factor: suppose hypothesis “A” says a data point is nearly impossible, and hypothesis “B” says the data point is quite likely. Then the existence of that one data point (by Bayes’ rule) should move you substantially toward believing hypothesis B (relative to A).
Example: you have had a rash on your arm for 10 years (with no variability). You buy some “rash cream” off of a shady website, and within 2 hours of applying it, the rash is gone. You can be confident the cream works because it’s otherwise highly unlikely for the rash to vanish.
Your comment makes sense. I think the problem goes even deeper.
Many nonfiction books project all of human experience onto a single axis and then ramble about that axis. In this case, the axis is “more” vs “less”. If you don’t understand what “more” and “less” are then this can be educational. But if you do know what “more” and “less” are then the important thing to understand is when should you apply this axis and when shouldn’t you. It is one half of a bravery debate.
Bravery debates are more about the listener than the facts. Whether you should subtract from your life depends on who you are.
He inhabited a sparse, unheated cell, its concrete walls radiating the late-fall chill. A wooden-plank tucket served as both bed and day couch, with a small stand alongside for perching texts to read—and little else. As befits a monk, the room was empty of any private belongings.
From the early-morning hours until late into the night, Khunu Lama would sit on that bed, a text always open in front of him. Whenever a visitor would pop in—and in the Tibetan world that could be at just about any time—he would invariably welcome them with a kindly gaze and warm words.
Khunu’s qualities—a loving attention to whoever came to see him—struck Dan as quite unlike, and far more positive than, the personality traits he had been studying for his degree in clinical psychology at Harvard. That training focused on negatives: neurotic patterns, overpowering burdensome feelings, and outright psychopathology.
Khunu, on the other hand, quietly exuded the better side of human nature. His humility, for instance, was fabled. The story goes that the abbot of the monastery, in recognition of Khunu’s spiritual status, offered him as living quarters a suite of rooms on the monastery’s top floor, with a monk to serve as an attendant. Khunu declined, preferring the simplicity of his small, bare monk’s cell.
―Altered Traits by Daniel Goleman and Richard J. Davidson
I don’t think Khunu needs to read about how he should subtract from his life.
The problem with Subtract is that the truth-value of its thesis depends on who is reading it. I prefer to read books with observer-independent (i.e. objective) truth values.
Yup, very well-put.
Your point about anecdotes got me thinking; an “extraordinary thesis” might be conceptualized as claiming that the distribution of data significantly shifted away from some “obvious” average. If so, showing the existence of a few data points greater than, say, 4 standard deviations from the “obvious” average actually can be strong evidence in its favor. However, the same is not true for a few examples ~2 standard deviations away. Maybe Klotz’s error is using anecdotes that aren’t far enough away from what intuitively makes sense.
Probably didn’t explain that very well, so here is a Tweet from Spencer Greenberg making the point:
Your comment makes sense. I think the problem goes even deeper.
Many nonfiction books project all of human experience onto a single axis and then ramble about that axis. In this case, the axis is “more” vs “less”. If you don’t understand what “more” and “less” are then this can be educational. But if you do know what “more” and “less” are then the important thing to understand is when should you apply this axis and when shouldn’t you. It is one half of a bravery debate.
Bravery debates are more about the listener than the facts. Whether you should subtract from your life depends on who you are.
I don’t think Khunu needs to read about how he should subtract from his life.
The problem with Subtract is that the truth-value of its thesis depends on who is reading it. I prefer to read books with observer-independent (i.e. objective) truth values.