I was on an airplane with my son who was just five years old at the time.

幾年前，

My son was so excited about being on this airplane with Mommy.

我和我兒子一起搭飛機， 他當時五歲。

He's looking all around and he's checking things out

對於能和媽咪一起搭飛機， 我兒子感到很興奮。

and he's checking people out.

他環視周圍，到處看看，

And he sees this man, and he says,

他也在打量其他人。

"Hey! That guy looks like Daddy!"

然後他看到一名男子，說：

And I look at the man,

「嘿！那個人看起來像爹地！」

and he didn't look anything at all like my husband,

我看向那名男子，

nothing at all.

他看起來完全不像我丈夫，

And so then I start looking around on the plane,

一點也不像。

and I notice this man was the only black guy on the plane.

所以我開始環視飛機內部，

And I thought,

我發現這名男子

"Alright.

是機上唯一的黑人男性。

I'm going to have to have a little talk with my son

我心想：「好。

about how not all black people look alike."

我要和我的兒子談一下，

My son, he lifts his head up, and he says to me,

讓他知道不是所有的黑人 看起來都一樣。」

"I hope he doesn't rob the plane."

我兒子抬起頭，對我說：

And I said, "What? What did you say?"

「我希望他不會在這架飛機上搶劫。」

And he says, "Well, I hope that man doesn't rob the plane."

我說：「什麼？你剛才說什麼？」

And I said, "Well, why would you say that?

他說：「我希望那個人 不會在這架飛機上搶劫。」

You know Daddy wouldn't rob a plane."

我說：「你為什麼會那樣說？

And he says, "Yeah, yeah, yeah, well, I know."

你知道爹地不可能會在飛機上搶劫。」

And I said, "Well, why would you say that?"

他說：「是啊，是啊，我知道。」

And he looked at me with this really sad face,

我說：「嗯，那你為什麼那樣說？」

and he says,

他看著我，表情很悲傷，

"I don't know why I said that.

他回答：

I don't know why I was thinking that."

「我不知道我為什麼那樣說。

We are living with such severe racial stratification

我不知道我為什麼那樣想。」

that even a five-year-old can tell us what's supposed to happen next,

我們生活在如此 嚴重的種族分層當中，

even with no evildoer,

就連五歲小孩也會告訴我們 接下來應該會發生什麼事，

even with no explicit hatred.

即使當下沒有作惡者，

This association between blackness and crime

也沒有明確的恨意。

made its way into the mind of my five-year-old.

黑人和犯罪之間的連結

It makes its way into all of our children,

進入了我五歲兒子的腦中。

into all of us.

它找到路進入了所有孩子的腦中，

Our minds are shaped by the racial disparities

我們所有人的腦中。

we see out in the world

我們在世界上所看到的種族差異

and the narratives that help us to make sense of the disparities we see:

形塑了我們的大腦，

"Those people are criminal."

這些說法協助我們理解 我們所看見的差異：

"Those people are violent."

「那些人是罪犯。」

"Those people are to be feared."

「那些人很暴力。」

When my research team brought people into our lab

「要害怕那些人。」

and exposed them to faces,

我的研究團隊把受試者 帶到我們的實驗室，

we found that exposure to black faces led them to see blurry images of guns

讓他們接觸不同的臉孔，

with greater clarity and speed.

我們發現，若他們 接觸的是黑人的臉孔，

Bias cannot only control what we see,

他們會更快、更清楚地

but where we look.

看出槍枝的模糊影像。

We found that prompting people to think of violent crime

偏見不僅會控制我們看見什麼，

can lead them to direct their eyes onto a black face

也會控制我們往哪裡看。

and away from a white face.

我們發現，若促使 受試者去想像暴力罪行，

Prompting police officers to think of capturing and shooting

他們會把視線轉向黑人臉孔，

and arresting

而非白人臉孔。

leads their eyes to settle on black faces, too.

若促使警察去想像 追捕、槍擊，以及逮捕，

Bias can infect every aspect of our criminal justice system.

也會導致他們把視線 落在黑人的臉孔上。

In a large data set of death-eligible defendants,

偏見會影響我們 刑事司法體系的每一方面。

we found that looking more black more than doubled their chances

在一個可判死刑之被告的 大型資料集中，

of receiving a death sentence --

我們發現，膚色比較黑的人

at least when their victims were white.

被判死刑的機會高達兩倍以上——

This effect is significant,

至少當被害人是白人時。

even though we controlled for the severity of the crime

這個影響非常顯著，

and the defendant's attractiveness.

就算我們控制了犯罪的嚴重程度

And no matter what we controlled for,

以及被告的吸引力也一樣。

we found that black people were punished

不論我們控制哪些變數，

in proportion to the blackness of their physical features:

我們都發現黑人受的懲罰

the more black,

和他們身體膚色有多黑成比例：

the more death-worthy.

越黑的人，

Bias can also influence how teachers discipline students.

越有可能被判死刑。

My colleagues and I have found that teachers express a desire

偏見也會影響老師如何懲戒學生。

to discipline a black middle school student more harshly

我和我同事發現，

than a white student

老師會想要對黑人中學生施以

for the same repeated infractions.

比白人學生更嚴厲的懲罰，

In a recent study,

即使他們重犯相同的錯誤。

we're finding that teachers treat black students as a group

在近期的一項研究中，我們發現

but white students as individuals.

老師會把黑人學生 當作一個群體來對待，

If, for example, one black student misbehaves

卻把白人學生當作 單獨的個體來對待。

and then a different black student misbehaves a few days later,

比如說，一個黑人學生 做出了不當的行為，

the teacher responds to that second black student

幾天後，有另一個黑人學生 也做了不當的行為，

as if he had misbehaved twice.

老師對第二個黑人學生的反應

It's as though the sins of one child

會像是他做了兩次不當的行為。

get piled onto the other.

就彷彿一個孩子的罪過

We create categories to make sense of the world,

可以疊加在另一個孩子身上。

to assert some control and coherence

為了理解世界，我們會創造分類，

to the stimuli that we're constantly being bombarded with.

為了稍微控制和整理

Categorization and the bias that it seeds

每天不斷轟炸我們的刺激物。

allow our brains to make judgments more quickly and efficiently,

分類以及它所種下的偏見

and we do this by instinctively relying on patterns

讓我們的大腦能快速 且有效地做出判斷，

that seem predictable.

我們會本能地依賴 看似可預測的模式。

Yet, just as the categories we create allow us to make quick decisions,

但，雖然我們創造的分類 能讓我們快速地做決策，

they also reinforce bias.

卻也會強化偏見。

So the very things that help us to see the world

所以，協助我們 理解世界的那些東西，

also can blind us to it.

也能矇蔽我們。

They render our choices effortless,

它們輕鬆地幫我們做出選擇，

friction-free.

完全沒有阻力。

Yet they exact a heavy toll.

但要付出的代價也很大。

So what can we do?

我們能怎麼做？

We are all vulnerable to bias,

我們大家都無法抵抗偏見，

but we don't act on bias all the time.

但我們不見得時時刻刻 都會依偏見行事。

There are certain conditions that can bring bias alive

有些條件會讓偏見活躍起來，

and other conditions that can muffle it.

有些則會抑制偏見。

Let me give you an example.

讓我舉個例子。

Many people are familiar with the tech company Nextdoor.

許多人都很熟悉 Nextdoor 這間科技公司。

So, their whole purpose is to create stronger, healthier, safer neighborhoods.

他們的目的是要創造出

And so they offer this online space

更堅強、更健康、更安全的鄰里。

where neighbors can gather and share information.

他們便提供這個線上空間，

Yet, Nextdoor soon found that they had a problem

讓鄰居可以聚集並分享資訊。

with racial profiling.

但，Nextdoor 很快就發現

In the typical case,

他們有種族歸納方面的問題。

people would look outside their window

在典型的情況中，

and see a black man in their otherwise white neighborhood

當居民看向窗外，

and make the snap judgment that he was up to no good,

看到他們的白人鄰里中有個黑人，

even when there was no evidence of criminal wrongdoing.

就會快速判斷這個黑人不懷好意，

In many ways, how we behave online

即便沒有任何證據顯示 他有打算要犯罪。

is a reflection of how we behave in the world.

我們在線上的許多行為方式

But what we don't want to do is create an easy-to-use system

都反映出我們在真實世界的行為。

that can amplify bias and deepen racial disparities,

但我們並不想要創造 一個操作簡易的系統

rather than dismantling them.

來放大偏見、加深種族差異，

So the cofounder of Nextdoor reached out to me and to others

而無法瓦解它們。

to try to figure out what to do.

Nextdoor 的共同創辦人

And they realized that to curb racial profiling on the platform,

便向我及他人求助， 希望能想出辦法來。

they were going to have to add friction;

他們知道，若要在平台上 抑制種族歸納，

that is, they were going to have to slow people down.

他們就得增加阻力；

So Nextdoor had a choice to make,

意即，他們要讓大家慢下來。

and against every impulse,

所以，Nextdoor 要做出選擇，

they decided to add friction.

他們決定不衝動行事，

And they did this by adding a simple checklist.

選擇加上阻力。

There were three items on it.

他們的做法是增添一個簡單的清單。

First, they asked users to pause

清單上有三個項目。

and think, "What was this person doing that made him suspicious?"

首先，他們請使用者暫停一下，

The category "black man" is not grounds for suspicion.

想想：「這個人做了什麼 讓他顯得可疑？」

Second, they asked users to describe the person's physical features,

「黑人」這個分類並不是 懷疑的基礎理由。

not simply their race and gender.

接著，他們請使用者描述 這個人的身體特徵，

Third, they realized that a lot of people

不單單是種族和性別。

didn't seem to know what racial profiling was,

第三，他們發現很多人

nor that they were engaging in it.

似乎不知道種族歸納是什麼，

So Nextdoor provided them with a definition

也不知道他們自己參與其中。

and told them that it was strictly prohibited.

所以 Nextdoor 為他們提供定義，

Most of you have seen those signs in airports

告訴他們種族歸納 是被嚴格禁止的。

and in metro stations, "If you see something, say something."

大部分的人在機場和地鐵站

Nextdoor tried modifying this.

都看過這類標牌：「如果 你看到了什麼事，說出來。」

"If you see something suspicious,

Nextdoor 把這個標牌改成：

say something specific."

「如果你看見了什麼可疑的事，

And using this strategy, by simply slowing people down,

明確地說出來。」

Nextdoor was able to curb racial profiling by 75 percent.

光是用這種讓大家慢下來的策略，

Now, people often will say to me,

Nextdoor 就讓 種族歸納減少了 75%。

"You can't add friction in every situation, in every context,

現在，大家通常會對我說：

and especially for people who make split-second decisions all the time."

「你不可能在所有情況、 所有情境中加上阻力，

But it turns out we can add friction

更不可能改變那些總是 瞬間做決定的人。」

to more situations than we think.

但結果發現，我們能添加阻力的情況

Working with the Oakland Police Department

比我們想像中的還要多。

in California,

和加州的奧克蘭警局合作之後，

I and a number of my colleagues were able to help the department

我和幾位同事能夠協助該警局

to reduce the number of stops they made

減少攔檢沒有犯下 任何嚴重罪行者的次數。

of people who were not committing any serious crimes.

我們的做法是要求警員

And we did this by pushing officers

在做每次攔檢時 都要先問自己一個問題：

to ask themselves a question before each and every stop they made:

「這次攔檢是有思考過 才決定要做的，

"Is this stop intelligence-led,

是或否？」

yes or no?"

換言之，

In other words,

我在攔檢之前是否有任何資訊，

do I have prior information to tie this particular person

能將這個人和特定的 犯罪連結在一起？

to a specific crime?

把那個問題加到警員 在攔檢時要填寫的表格上，

By adding that question

就會讓警員慢下來，暫停一下，

to the form officers complete during a stop,

他們會想：「我為什麼 會想要把這個人攔下來？」

they slow down, they pause,

在 2017 年，我們把 那個思考問題納入表格之前，

they think, "Why am I considering pulling this person over?"

全市的警員進行了 大約 32,000 次攔檢。

In 2017, before we added that intelligence-led question to the form,

隔年，加上了那個問題之後，

officers made about 32,000 stops across the city.

數字減少至 19,000 次攔檢。

In that next year, with the addition of this question,

光是對非裔美國人 所做的攔檢就減少了 43%。

that fell to 19,000 stops.

且減少對黑人的攔檢次數 並沒有讓城市變得比較危險。

African-American stops alone fell by 43 percent.

事實上，犯罪率持續下降，

And stopping fewer black people did not make the city any more dangerous.

該城市變得讓每個人都更安全。

In fact, the crime rate continued to fall,

一個解決方案是減少 不必要的攔檢數目。

and the city became safer for everybody.

另一個解決方案則是必須攔檢時，

So one solution can come from reducing the number of unnecessary stops.

改善攔檢的品質。

Another can come from improving the quality of the stops

此時科技就能派上用場。

officers do make.

我們都知道喬治．佛洛伊德的死亡，

And technology can help us here.

因為當時試圖協助他的人

We all know about George Floyd's death,

用手機攝影機拍下

because those who tried to come to his aid held cell phone cameras

這段和警方可怕、致命的遭遇。

to record that horrific, fatal encounter with the police.

我們有各種科技， 但我們沒有善用。

But we have all sorts of technology that we're not putting to good use.

全國各地的警局

Police departments across the country

現在都被要求在身上戴攝影機，

are now required to wear body-worn cameras

這麼一來，我們會記錄到的 不僅是最可怕極端的遭遇，

so we have recordings of not only the most extreme and horrific encounters

還會記錄日常的互動。

but of everyday interactions.

我們和史丹佛大學的 一個跨領域團隊合作，

With an interdisciplinary team at Stanford,

開始使用機器學習技術

we've begun to use machine learning techniques

來分析大量相遇的情境。

to analyze large numbers of encounters.

目的是要更了解在例行的 交通攔檢時會發生什麼事。

This is to better understand what happens in routine traffic stops.

我們發現，

What we found was that

即使警員的行為表現非常專業，

even when police officers are behaving professionally,

比起白人司機，他們對黑人司機的 說話方式仍然比較不尊重。

they speak to black drivers less respectfully than white drivers.

事實上，單單從警員的用詞，

In fact, from the words officers use alone,

我們就能預測出他們是在 和黑人司機或白人司機說話。

we could predict whether they were talking to a black driver or a white driver.

問題是，這些攝影機 所拍攝的大部分影片

The problem is that the vast majority of the footage from these cameras

並沒有被警局用來 了解街上發生的狀況

is not used by police departments

或用來訓練警員。

to understand what's going on on the street

那很可惜。

or to train officers.

例行的攔檢怎麼會 變成致命的遭遇？

And that's a shame.

在喬治．佛洛伊德的案例中 這是怎麼發生的？

How does a routine stop turn into a deadly encounter?

在其他的案例中呢？

How did this happen in George Floyd's case?

我的長子 16 歲時，

How did it happen in others?

他發現當白人看著他時，

When my eldest son was 16 years old,

他們會感到恐懼。

he discovered that when white people look at him,

他說，最糟的時候是在電梯裡。

they feel fear.

當電梯門關上，

Elevators are the worst, he said.

大家被困在這個小小的空間裡，

When those doors close,

和他們認為危險的人困在一起。

people are trapped in this tiny space

我兒子能感受到他們的不適，

with someone they have been taught to associate with danger.

並用微笑讓他們感到自在些，

My son senses their discomfort,

降低他們的恐懼。

and he smiles to put them at ease,

當他說話時，

to calm their fears.

他們的身體放鬆了。

When he speaks,

他們比較能呼吸。

their bodies relax.

他們喜歡他的抑揚頓挫、

They breathe easier.

他的發音、他的用字遣詞。

They take pleasure in his cadence,

他聽起來就像他們的一份子。

his diction, his word choice.

我以前認為我兒子 和他爸爸一樣天生外向。

He sounds like one of them.

但，在那一刻， 我從那段談話中了解到，

I used to think that my son was a natural extrovert like his father.

他的微笑並不表示他想要

But I realized at that moment, in that conversation,

和那些本來會是陌生人的人連結。

that his smile was not a sign that he wanted to connect

他的微笑是個護身符， 用來保護他自己，

with would-be strangers.

是他搭電梯搭了數千次 而磨練出來的生存技能。

It was a talisman he used to protect himself,

他的膚色會產生緊張感， 讓他自己有生命危險，

a survival skill he had honed over thousands of elevator rides.

他便學習去調解這樣的緊張感。

He was learning to accommodate the tension that his skin color generated

我們知道大腦天生就會有偏見，

and that put his own life at risk.

打破那偏見的一種方式 就是先暫停下來做反思，

We know that the brain is wired for bias,

想想我們的假設有什麼依據。

and one way to interrupt that bias is to pause and to reflect

我們得要問問自己：

on the evidence of our assumptions.

我們帶著什麼假設踏進電梯？

So we need to ask ourselves:

或上飛機？

What assumptions do we bring when we step onto an elevator?

我們要如何察覺到 自己無意識的偏見？

Or an airplane?

那些假設能保護誰的安全？

How do we make ourselves aware of our own unconscious bias?

那些假設讓誰有危險？

Who do those assumptions keep safe?

若我們不去問這些問題，

Who do they put at risk?

不去堅持我們的學校、法庭、警局，

Until we ask these questions

及其他機構都要問這些問題，

and insist that our schools and our courts and our police departments