First let me hedge that I have not read this book, White Fragility. The exposures I’ve had to this kind of literature drove me to stay far away. The following critiques may not be about this specific book nor author, DiAngelo. Perhaps DiAngelo-adjacent, though I suspect also directly DiAngelo for some points.
1) This kind of literature has serious problems with semantic ambiguity. You can make anything meaningless by saying it’s everything, and “racism” is no exception. I imagine myself trying to predict the degree of “fragility” or “racism” in members of the population, or changes in “fragility” or “racism” over time, as the author uses these terms. It seems thoroughly unresolvable and nebulous. I don’t know how you would make these forecasting questions without them resolving ambiguously all the time. That is a very negative fact about how informative the content is.
2) What appears to be an abysmal performance at predicting changes in actual harm or wellbeing, in blacks or other groups. I don’t know of explicit proper predictions, so this is hard to grade. But what I remember seeing doesn’t look good. Again I don’t know what DiAngelo in particular has predicted before, but I’ll go through some related examples from people who seem to hold similar attitudes.
One is the prediction that diversity training would causally increase the relative workplace status of blacks. This prediction seems false.
Another is the rampant idea that intelligence test screening causally reduces the number of gifted blacks in gifted programs, thinking the tests are unfairly biased against blacks. This prediction is false, and using intelligence screening actually makes it easier to put gifted blacks where they belong.
Another is the prediction that reducing police presence would reduce harm for blacks. This seems highly unlikely to me, and I would predict harm toward blacks goes up in areas that get rid of the cops. Not down.
Or the idea that you can reduce harm by training cops to not be differentially-trigger-happy towards blacks. This can’t be correct because they’re not actually more trigger-happy against blacks. At least, not in the U.S.
Or the prediction that training cops to avoid shooting blacks could make a difference to the average lifespan of blacks. This is impossible—out of 42 million blacks in the U.S., a little over 200 per year are shot to death by cops. For context that’s more than the number that die from lightning strikes, but less than the number that die from drowning.
Or that rent control reduces harm for minorities. It doesn’t, actually it makes things worse.
In a futarchy where there is an incentive to predict the effects of policies, I expect a poor performance from the kinds of policies these people propose. This despite the loud advertisements that they know how to improve wellbeing for blacks. And with such confidence that they should be given government subsidy, and ostracizing power over those who disagree. These people expect me to think they know how to reduce harm for blacks. They think their ideas are slam-dunks, and it’s absurd. I can’t even tell if they outperform random chance at predicting changes in harm in blacks. That is severely damning of their judgment.
3) The framing tends to be manipulative. Whether a white person condemns racism or doesn’t condemn racism, both are described as instigating racism. And of course, any disagreement is evidence that you’re a bigot. But so is agreement—must be some treachery by defusing the conflict today, so you can keep oppressing blacks tomorrow. It’s a bizarre mixture of voodoo-cursed-thoughts and original sin, and it’s unacceptable to me.
In summary, this kind of literature isn’t remotely compatible with my epistemology. Nor how I think discourse about harm and wellbeing should be conducted. In fact it is bitterly exemplary of how I would not want it, and the pressure to conform is not okay with me.
Or the prediction that training cops to avoid shooting blacks could make a difference to the average lifespan of blacks. This is impossible—out of 42 million blacks in the U.S., a little over 200 per year are shot to death by cops. For context that’s more than the number that die from lightning strikes, but less than the number that die from drowning.
Concretely:
200 deaths/year*(75 years/lifetime)/42 million lifetimes)*40 years lost *(365 days/years) ~= 5.2 days/lifetime, so 5 days is the average lifetime lost for black people compared to if you get rid of all police shootings and there are no other secondary effects.
Realistically getting rid of 100% of police shootings is unrealistic, but 20%-50% (or extending average black lifetimes by 1-3 days) doesn’t seem crazy to me.
I multiplied these numbers out because the dimensional analysis for yearly death rate and number of total people alive is pretty confusing unless you have an intuition for this stuff (which I at least don’t have enough of), can imagine people walking away from just the raw numbers thinking the expected per capita loss is closer to hours or closer to weeks.
First let me hedge that I have not read this book, White Fragility. The exposures I’ve had to this kind of literature drove me to stay far away. The following critiques may not be about this specific book nor author, DiAngelo. Perhaps DiAngelo-adjacent, though I suspect also directly DiAngelo for some points.
1) This kind of literature has serious problems with semantic ambiguity. You can make anything meaningless by saying it’s everything, and “racism” is no exception. I imagine myself trying to predict the degree of “fragility” or “racism” in members of the population, or changes in “fragility” or “racism” over time, as the author uses these terms. It seems thoroughly unresolvable and nebulous. I don’t know how you would make these forecasting questions without them resolving ambiguously all the time. That is a very negative fact about how informative the content is.
2) What appears to be an abysmal performance at predicting changes in actual harm or wellbeing, in blacks or other groups. I don’t know of explicit proper predictions, so this is hard to grade. But what I remember seeing doesn’t look good. Again I don’t know what DiAngelo in particular has predicted before, but I’ll go through some related examples from people who seem to hold similar attitudes.
One is the prediction that diversity training would causally increase the relative workplace status of blacks. This prediction seems false.
Another is the rampant idea that intelligence test screening causally reduces the number of gifted blacks in gifted programs, thinking the tests are unfairly biased against blacks. This prediction is false, and using intelligence screening actually makes it easier to put gifted blacks where they belong.
Another is the prediction that reducing police presence would reduce harm for blacks. This seems highly unlikely to me, and I would predict harm toward blacks goes up in areas that get rid of the cops. Not down.
Or the idea that you can reduce harm by training cops to not be differentially-trigger-happy towards blacks. This can’t be correct because they’re not actually more trigger-happy against blacks. At least, not in the U.S.
Or the prediction that training cops to avoid shooting blacks could make a difference to the average lifespan of blacks. This is impossible—out of 42 million blacks in the U.S., a little over 200 per year are shot to death by cops. For context that’s more than the number that die from lightning strikes, but less than the number that die from drowning.
Or that rent control reduces harm for minorities. It doesn’t, actually it makes things worse.
In a futarchy where there is an incentive to predict the effects of policies, I expect a poor performance from the kinds of policies these people propose. This despite the loud advertisements that they know how to improve wellbeing for blacks. And with such confidence that they should be given government subsidy, and ostracizing power over those who disagree. These people expect me to think they know how to reduce harm for blacks. They think their ideas are slam-dunks, and it’s absurd. I can’t even tell if they outperform random chance at predicting changes in harm in blacks. That is severely damning of their judgment.
3) The framing tends to be manipulative. Whether a white person condemns racism or doesn’t condemn racism, both are described as instigating racism. And of course, any disagreement is evidence that you’re a bigot. But so is agreement—must be some treachery by defusing the conflict today, so you can keep oppressing blacks tomorrow. It’s a bizarre mixture of voodoo-cursed-thoughts and original sin, and it’s unacceptable to me.
In summary, this kind of literature isn’t remotely compatible with my epistemology. Nor how I think discourse about harm and wellbeing should be conducted. In fact it is bitterly exemplary of how I would not want it, and the pressure to conform is not okay with me.
Concretely:
200 deaths/year*(75 years/lifetime)/42 million lifetimes)*40 years lost *(365 days/years) ~= 5.2 days/lifetime, so 5 days is the average lifetime lost for black people compared to if you get rid of all police shootings and there are no other secondary effects.
Realistically getting rid of 100% of police shootings is unrealistic, but 20%-50% (or extending average black lifetimes by 1-3 days) doesn’t seem crazy to me.
I multiplied these numbers out because the dimensional analysis for yearly death rate and number of total people alive is pretty confusing unless you have an intuition for this stuff (which I at least don’t have enough of), can imagine people walking away from just the raw numbers thinking the expected per capita loss is closer to hours or closer to weeks.