Critical race theory

if you know why your rationale/evidentiary burden isn't met in advance, it's odd to keep making the case despite that

i'm sure we know ways the theory could be tested at scale
 
if you know why your rationale/evidentiary burden isn't met in advance, it's odd to keep making the case despite that

i'm sure we know ways the theory could be tested at scale

Under the definitions you insist on working under, I'm not sure we do.

As you insist that racism must be a result of a personal action (or overt policy), this means at levels of detail where an outcome is the summed result of five actions (and policy is covert/unknowable), we can never know enough to determine your standard of racism. It would help if we had a labor market in a pocket universe which we had complete informational access upon.

Your attitude suggests you consider this a feature, not a bug. It looks about as good faith as requiring an extremely strict standard of evidence for if a patient died of COVID, or died with COVID.
 
Under the definitions you insist on working under, I'm not sure we do.

As you insist that racism must be a result of a personal action (or overt policy), this means at levels of detail where an outcome is the summed result of five actions (and policy is covert/unknowable), we can never know enough to determine your standard of racism. It would help if we had a labor market in a pocket universe which we had complete informational access upon.

Your attitude suggests you consider this a feature, not a bug. It looks about as good faith as requiring an extremely strict standard of evidence for if a patient died of COVID, or died with COVID.

What's really funny about this is that literal Jim Crow laws preventing African-Americans from voting would fail to meet this evidentiary standard because, due to the passage of the 15th Amendment to the US Constitution, states could not just directly ban African-Americans from voting and had to come up with facially neutral pretexts (such as the poll tax/grandfather clause, a combo as iconic to America as a cheeseburger with fries).
 
Crunched for time, but "they" being the ones that shut down when they hear "CRT." It's a signpost and trust applies. Since the wondering of why was directly related to the topic not believed in.

It's also why you can agree with the basic premises as they hit the road but still criticize the frame itself for being... heavy. Let's use "heavy."


How do we feed the vegans, Farm Boy? https://teddit.net/r/BestofRedditor...flipping_out_on_my_fiance_for_cancelling_all/
 
Well, I'm not going to have sympathy. They're merging lives, if they can't share food, that bodes ill.

I guess I'm too flipping around to sort it out. I'm not making the connection you're aiming for.
 
Like obviously the couple’s answer is holy smokes you cannot get married.

But say the wedding has to happen. These people are adamant your vegan food is trying to shame them and exclude them, that you are taking from them, and it’s no big deal to just equally partake in their salad option. But you’re the bride and your family is vegan. You have to feed them. How can you reach these people?
 
If you can't when you're the bride: you don't.

If they're not total sacks of ****, weeping that you can't invite your family to your own wedding would work. But, that seems to be a stretch, considering.

Either way, fighting and public shame as a near-first of first approach doesn't seem effective. Generally.
 
Last edited:
What's really funny about this is that literal Jim Crow laws preventing African-Americans from voting would fail to meet this evidentiary standard because, due to the passage of the 15th Amendment to the US Constitution, states could not just directly ban African-Americans from voting and had to come up with facially neutral pretexts (such as the poll tax/grandfather clause, a combo as iconic to America as a cheeseburger with fries).
Don't forget the "separate but equal" rule, which enabled whites to segregate blacks from their white society and their seal h. Which is what triggered whites in Oklahoma in 1922 when they slaughtered at least 300 people of color and burned dozens of blocks of residences andbusiness. Greenville, aka "The Black Wall Street", was never rebuilt, most of the land was scooped up by whites when white banks refused to loan money to rebuild. Aso, not one person was ever arrested, much less charged with a crime, in connection with this genocidal act.

But sure, no systematic oppression of blacks.
that isn't "systemic hidden racism that doesn't track to reality", though. that's direct evidence of discrimination. the kind of evidence i ask for broadly, if you want to make broad assertions. the kind that correlations should drive you to look for and root out when you find it (and look for other explanations if you don't).

there is an enormous difference between "this particular industry is > 80% men, and that's a problem because reasons" vs "our model predicts bias against x, then we test for that bias, and observe evidence consistent with the prediction". though i would still be interested in what type of position is being applied for and the callback rate for the others (white with degrees/no criminal record, black with criminal history/no degree), as it informs the bias further.
You're talking about the private zector, not government.
The private sector is only bound by laws emanating from the government.
If the laws allow discrimination, the private sector will take advantage.
If the laws prohibit discrimination the private sector will have to work harder to discriminate and will face penalties for doing so.
So if a set of laws, lets call them Tim Titwillow laws, allows discrimination and segregation against people with blonde hair, then blondes won't have more fun.
 
Last edited:
Btw equilibrium is the optimal point of the prisoner's dilemma. Both semantically (aka what the words mean) and in terms of outcome.
Sorry, was meaning to say optimal sum of reward isn't the equilibrium of that dilemma. I think there are dynamic systems that exist no equilibrium, one of such is the solar system, which is harmonic.
The labor market might be more "efficient" than 150 years ago but that doesn't mean it has arbitraged to equal wages.
Maybe it is really efficient nowadays or maybe it is not. Isn't waiting for market fixing itself a simpler solution that enacts unnatural laws based on some controversial theories that hasn't reached a consensus?
If most of society already accepted CRT, there shouldn't be any visible difference in opportunity in $ bracket for any "underprivileged" race. (It is not per $ because opportunity and $ should not have linearity, it would be like, you know, something like opportunity = a*2^$).
Making statistics not considering $ of each race is bs because $ per capita in each race vastly different. We all know and accept people with higher $ having more opportunity, right?
-
Let's imagine that people with those particular features, are harder to get certain jobs. Well the above thing is true, everyone knows being ugly decrease your wage in some jobs.
Most importantly, ugliness is subjective, and there are evidences that people are attracted to people that look like them.
Hiring the people in same biological "race" might help something like advertisements and customer service gaining more attraction passively, which might have some good effect somehow.
The demand of people of a particular dominant race (race with highest purchasing power, not race with most people) increases the price to hire those in those particular field.
That's very natural and there are absolutely nothing wrong with that. It's our specie's bias.
And making statistics without excluding all of those particular kinds of jobs, is wrong too.
-
There are many hidden factors that we don't know, so instead of trying to do social engineering, let nature balance it out.
Let the free market choose what to do and trust its decision.
Imagine a world that needs laws to tell people stop racism, horrible.
-
People are blinded by race when making statistics, while $ shows a greater correlation.
Equal wage is morally wrong.
-
How to make right statistics on those social problems? Idk, ask god.
 
Sorry, was meaning to say optimal sum of reward isn't the equilibrium of that dilemma. I think there are dynamic systems that exist no equilibrium, one of such is the solar system, which is harmonic.

Maybe it is really efficient nowadays or maybe it is not. Isn't waiting for market fixing itself a simpler solution that enacts unnatural laws based on some controversial theories that hasn't reached a consensus?
If most of society already accepted CRT, there shouldn't be any visible difference in opportunity in $ bracket for any "underprivileged" race. (It is not per $ because opportunity and $ should not have linearity, it would be like, you know, something like opportunity = a*2^$).
Making statistics not considering $ of each race is bs because $ per capita in each race vastly different. We all know and accept people with higher $ having more opportunity, right?
-
Let's imagine that people with those particular features, are harder to get certain jobs. Well the above thing is true, everyone knows being ugly decrease your wage in some jobs.
Most importantly, ugliness is subjective, and there are evidences that people are attracted to people that look like them.
Hiring the people in same biological "race" might help something like advertisements and customer service gaining more attraction passively, which might have some good effect somehow.
The demand of people of a particular dominant race (race with highest purchasing power, not race with most people) increases the price to hire those in those particular field.
That's very natural and there are absolutely nothing wrong with that. It's our specie's bias.
And making statistics without excluding all of those particular kinds of jobs, is wrong too.
-
There are many hidden factors that we don't know, so instead of trying to do social engineering, let nature balance it out.
Let the free market choose what to do and trust its decision.
Imagine a world that needs laws to tell people stop racism, horrible.
-
People are blinded by race when making statistics, while $ shows a greater correlation.
Equal wage is morally wrong.
-
How to make right statistics on those social problems? Idk, ask god.

This is a really quite incredible amount of effort to justify doing nothing. Also, the claim that the free market is the default and "natural" position.
 
If you can't when you're the bride: you don't.

If they're not total sacks of ****, weeping that you can't invite your family to your own wedding would work. But, that seems to be a stretch, considering.

Either way, fighting and public shame as a near-first of first approach doesn't seem effective. Generally.
Sorry, was meaning to say optimal sum of reward isn't the equilibrium of that dilemma. I think there are dynamic systems that exist no equilibrium, one of such is the solar system, which is harmonic.

Maybe it is really efficient nowadays or maybe it is not. Isn't waiting for market fixing itself a simpler solution that enacts unnatural laws based on some controversial theories that hasn't reached a consensus?
If most of society already accepted CRT, there shouldn't be any visible difference in opportunity in $ bracket for any "underprivileged" race. (It is not per $ because opportunity and $ should not have linearity, it would be like, you know, something like opportunity = a*2^$).
Making statistics not considering $ of each race is bs because $ per capita in each race vastly different. We all know and accept people with higher $ having more opportunity, right?
-
Let's imagine that people with those particular features, are harder to get certain jobs. Well the above thing is true, everyone knows being ugly decrease your wage in some jobs.
Most importantly, ugliness is subjective, and there are evidences that people are attracted to people that look like them.
Hiring the people in same biological "race" might help something like advertisements and customer service gaining more attraction passively, which might have some good effect somehow.
The demand of people of a particular dominant race (race with highest purchasing power, not race with most people) increases the price to hire those in those particular field.
That's very natural and there are absolutely nothing wrong with that. It's our specie's bias.
And making statistics without excluding all of those particular kinds of jobs, is wrong too.
-
There are many hidden factors that we don't know, so instead of trying to do social engineering, let nature balance it out.
Let the free market choose what to do and trust its decision.
Imagine a world that needs laws to tell people stop racism, horrible.
-
People are blinded by race when making statistics, while $ shows a greater correlation.
Equal wage is morally wrong.
-
How to make right statistics on those social problems? Idk, ask god.
The sum of the reward in a multi game, multi partner version:

Prisoner's DilemmaCooperateDefect
Cooperate3 + 3 = 61 + 4 = 5
Defect4 + 1 = 52 + 2 = 4
 
The sum of the reward in a multi game, multi partner version:

Prisoner's DilemmaCooperateDefect
Cooperate3 + 3 = 61 + 4 = 5
Defect4 + 1 = 52 + 2 = 4
yeah but equilibrium is when both defect, yields the sum of reward of 4, is that correct, which is not the most optimal sum of reward but the optimal for each player.
 
So there's no market here, which makes "equilibrium" a strange concept, or rather, an imperfect fit. But there is an inferred metagame (a term economics is slow to adopt).

So assuming that your play history with your one partner is known by your other partner, and that they factor it, and that they aren't martyrs for the sake of total economy:

In a single round 1v1, the individual is incentivized to defect for a maximum possible outcome (2 and 4 points vs 1 and 3).

In a multi round 1v1, you would cooperate until the last round, but then logically the second to last round, until the logic is no cooperation all the way back. Something something going against a Sicilian.

In a multi round free for all, outside cooperators are kicking your ass because your dumbass was stuck in a defection loop trying to cheat single partners. But if you were a cooperate-only then you will get taken advantage of by other defectors. So the plan is to cooperate but to not get suckered. The most basic optimum is tit-for-tat with expected charity: you start by cooperating and then you copy your opponents' previous move. Fellow cooperators get rewarded, fellow defectors get punished, and weirdos get their weirdness back for a mix.
 
Fighting can't always yield results. Arguing, especially where other people can hear or read it, is even less likely to yield results.

But a good story, sometimes that can move the world. That's the most appropriate takeaway from your wedding dilemma. The problems with the food selection are pride and fear of shame. A public confrontation in front of the wrong people amplifies that. Like yes, "holy crap those people probably shouldn't marry" went from possibly true(young people burn hot and not always wisely) to being almost certainly true with the confrontation. If quiet inflexibility on the love for one's own family having a meal they can eat without being looked down upon would have remained a division point, then yes - the "don't marry" becomes almost certain.
 
What's really funny about this is that literal Jim Crow laws preventing African-Americans from voting would fail to meet this evidentiary standard because, due to the passage of the 15th Amendment to the US Constitution, states could not just directly ban African-Americans from voting and had to come up with facially neutral pretexts (such as the poll tax/grandfather clause, a combo as iconic to America as a cheeseburger with fries).

The biggest one was the literacy test. It was an out and out fraud.
 
Fighting can't always yield results. Arguing, especially where other people can hear or read it, is even less likely to yield results.
Depends on what you mean by "results". There are common contexts where convincing the other party is very, very unlikely, so aiming for that result is unrealistic. But the audience might reap benefits - some could be convinced, and some could be heartened that you're taking their side. Especially important when a majority is oppressing a minority.
 
Depends on what you mean by "results". There are common contexts where convincing the other party is very, very unlikely, so aiming for that result is unrealistic. But the audience might reap benefits - some could be convinced, and some could be heartened that you're taking their side. Especially important when a majority is oppressing a minority.

Of course. There is a sort of vanity in that which is not always wrong. But it also highlights a fundamental weakness of the frame pivot from discrimination to privilege. If the story is told in the lens of discrimination, you have a hero(es). They have companions, and allies, and enemies, and the non-involved. The story, as it passes down, invites choosing the side of good. When the privilege is inborn, and the history immutable because the numberlords need to have a nice simple easy-virtue to knock down targets, it doesn't. The history is written, those who were wrong are wrong, and the strength of potential allies is ill-gotten.

The frame is terrible and heavy*. And the people imperfect and fickle. It doesn't bring out the best. Or, at the very least, its a degradation of what could be better.

*and yes, misused by malice. Of course.
 
It doesn't bring out the best. Or, at the very least, its a degradation of what could be better.

Agree.

It's by design. Corporations push it because if the black workers walk out of a diversity training session distrustful of their white colleagues, and the white workers walk out angry at their black colleagues, it's a huge win. The workforce has been divided and conquered, and will be much less likely to unionize or engage in any other collective action for their own benefit. They'll be too busy punching laterally to punch up.
 
Agree.

It's by design. Corporations push it because if the black workers walk out of a diversity training session distrustful of their white colleagues, and the white workers walk out angry at their black colleagues, it's a huge win. The workforce has been divided and conquered, and will be much less likely to unionize or engage in any other collective action for their own benefit. They'll be too busy punching laterally to punch up.
I literally haven't seen this happen in nearly a decade of doing them. Have you?

Like, it's a mild administrative inconvenience. It's useless, sure. It doesn't do much other than tick boxes. But that's not what you're saying here. That sounds like more "culture war" nonsense.
 
I literally haven't seen this happen in nearly a decade of doing them. Have you?

Like, it's a mild administrative inconvenience. It's useless, sure. It doesn't do much other than tick boxes. But that's not what you're saying here. That sounds like more "culture war" nonsense.

Robin DiAngelo is getting rich off of somebody.

Someone who wanted to make a good faith effort to build bridges would bring in Chloe Valdary instead.
 
Back
Top Bottom