The Association Mechanism Is Stereotyping
We are studying how we can make our understandings accurate, but Earthians don't seem to care about such a thing. They are happy to skew their understandings for their advantage.
Hmm, we have a problem . . .
Yes, it's a problem!
I'm talking about you.
Huh?
You are stereotyping things, which is the problem.
Am I?
All the Earthians aren't the same. When you say, "Earthians don't . . .", whom are you talking about?
Ah . . .
Such statements are meaningless, or at least ambiguous. If you claim something about all the Earthians, state so explicitly.
. . . Certainly, it was a stereotype. Some Earthians aren't like that, or at least we don't have any proof that shows that all the Earthians are like that.
The fact that you who are trying to remedy biases tend to be easily trapped in stereotyping shows how deep the mechanism of stereotyping is rooted in human mentality.
That's true. I thought I was careful not to do stereotyping, but stereotyping slips into my thinking process. How does that happen?
The intuition is basically an association factory, and associations are stereotypes. As the intuition incessantly works regardless of our intentions, it's quite difficult to check all the stereotypes.
The association just connects two concepts, for example, woman -> weak. So, when I see any woman, according to the association, she is weak: that's stereotyping. If we fully entrust judgments to the intuition, stereotypes will just flow out of the intuition, as raw waste fluid flow out from a factory.
Yes. We probably can't stop the factory, but we have to try to clean the waste fluid as much as possible.
We can't completely clean the waste fluid, can we?
I don't think we can. However, a problem is to think, "As we can't completely clean the waste fluid anyway, it doesn't matter how polluted it is." It isn't all or nothing. The cleaner, the better.
We Have to Think Each Group as a Set with Diverse Members
The association mechanism is problematic because it doesn't understand the group as a set in which the members are different from each other.
When we think of an attribute, for example strength, usually, each member of a group has a different value, and the members of the group are distributed along the attribute value axis. That's the accurate understanding.
A typical distribution is the normal distribution. The strengths of women, for example, will be in a distribution that resembles the normal distribution. Such a statement as "Women are weak" is ambiguous. Is the average strength weak? Well, the problem is that someone who claims that statement tends to do a deduction like this: "Women are weak. You are a woman. So, you are weak." That's absurd. Why can one conclude that a woman is weak because the average strength of women is weak?
Yes, absurd. In the normal distribution, there are strong women even if the average strength of women is weaker than men's. The stronger, the fewer, but there are women who are stronger than most of men.
Some people would say, "My statement was just rhetoric! I understand that there are some strong women," but, the problem is, their deductions show that they don't understand well.
That may have been just rhetoric, but such rhetoric tricks both the one who said it and people who heard it.
By the way, there are some cases in which all the members have really the same attribute value.
Of course, there are, but still it's a kind of distribution, and we can reach accurate understandings by examining the distribution: the distribution just happened to be a one-value distribution. On the other hand, by the association mechanism, we can't reach accurate understandings because it precludes the possibility of the existence of distributions from the beginning.
When we think of a distribution, first, we should suppose the distribution as a normal distribution, right?
I think, that's a sound strategy. In the nature world, the normal distribution is the normal. Besides, as the intuition loves the one-value distribution, supposing the one-value distribution as the default is detrimental: we have the undesirable incentive to stick to the one-value distribution against evidences. It will be wise to reserve the one-value distribution as the last option.
Why Can We Stick to Obviously-False Stereotypes?
A wonder is that we can stick to obviously-false stereotypes. For example, we know as knowledge that female wrestlers exist and female great scientists exist, and still can say like "Women are weak" and "Women are incompetent". . . . How can that happen? What are we thinking?
Ah, that's a good question. I guess, the thinking paradigm we are in limits our field of vision. As the strategy of the association paradigm is to make one typical model member represent a group, the diversity among the members of the group is just out of its concern: such a concept as diversity doesn't exist for it. Female great scientists are just out of our awareness.
Ah-ha, a limitation of human thinking mechanism is that we aren't aware of the whole of our knowledge at the same time. Even if we have a piece of knowledge stored in our brain somewhere, as far as we don't recall that piece of knowledge, we can nonchalantly make a deduction that totally contradicts that piece of knowledge.
That's an important characteristic of human thinking mechanism. We can't expect that we can judge things based on our knowledge just because we have the knowledge: we have to recall the knowledge.
The problem of the association paradigm is that it dissuades us from looking through our knowledge.
That enables quick judgments by removing the necessity for looking through the whole of our knowledge. That's its purpose. That's its essence.
So, what can we do? We should always imagine a bell curve when we think of a group.
As the intuition inevitably does stereotyping, we have to counteract stereotypes by intentionally diversifying the distributions.
