Should you trust unanimous decisions? - Derek Abbott

譯者: Helen Lin 審譯者: Max Chern

Imagine a police lineup where ten witnesses

想像一個「列隊指認」， 十位目擊者

are asked to identify a bank robber they glimpsed fleeing the crime scene.

被要求指認在現場瞥見 逃跑的銀行搶劫犯。

If six of them pick out the same person,

如果其中六位指認同一人，

there's a good chance that's the real culprit,

那他很可能是真的罪犯，

and if all ten make the same choice,

如果十人的指認都相同，

you might think the case is rock solid,

你可能想這情況罪證確鑿，

but you'd be wrong.

但你錯了！

For most of us, this sounds pretty strange.

我們多數會覺得這聽起來很奇怪，

After all, much of our society relies on majority vote and consensus,

畢竟，我們社會大多 依賴 多數表決與共識，

whether it's politics,

無論是政治、商業或娛樂活動。

business,

or entertainment.

So it's natural to think that more consensus is a good thing.

所以視 ‘越多共識是件好事’ 是理所當然的。

And up until a certain point, it usually is.

到某個程度，它通常是這樣的。

But sometimes, the closer you start to get to total agreement,

但有時，當開始越靠近完全一致時，

the less reliable the result becomes.

結果變得越不可靠，

This is called the paradox of unanimity.

這稱為「一致性悖論」。

The key to understanding this apparent paradox

了解這明顯矛盾論點的關鍵

is in considering the overall level of uncertainty

在於考量整體不確定性，

involved in the type of situation you're dealing with.

它涉及你正在處理的情況類型。

If we asked witnesses to identify the apple in this lineup, for example,

例如，我們要求目擊者 指認這個列隊中的蘋果，

we shouldn't be surprised by a unanimous verdict.

對有一致性的判斷 是無需驚訝的。

But in cases where we have reason to expect some natural variance,

但有理由預料會有些自然差異的存在下，

we should also expect varied distribution.

我們也應預期會有不同的分佈。

If you toss a coin one hundred times,

如果你擲一個硬幣一百次，

you would expect to get heads somewhere around 50% of the time.

你會預期得到人頭的次數 約在 50% 左右，

But if your results started to approach 100% heads,

但若結果開始趨近 100% 的人頭，

you'd suspect that something was wrong,

你會懷疑有什麼不對勁了，

not with your individual flips,

問題不在於你每次的拋擲，

but with the coin itself.

而是硬幣本身。

Of course, suspect identifications aren't as random as coin tosses,

當然，嫌疑犯的指認 不同於擲硬幣的隨機性，

but they're not as clear cut as telling apples from bananas, either.

但也不像分辨蘋果與香蕉那樣明確。

In fact, a 1994 study found that up to 48% of witnesses

事實上，一個 1994 年的研究發現 高達 48% 的目擊者

tend to pick the wrong person out of a lineup,

往往會在列隊中指認錯人，

even when many are confident in their choice.

即便許多人很確信自己的指認。

Memory based on short glimpses can be unreliable,

短暫一瞥的記憶 可能是靠不住的，

and we often overestimate our own accuracy.

而且我們常高估自己的準確度。

Knowing all this,

知道這些之後，

a unanimous identification starts to seem less like certain guilt,

一致性的指認開始看起來 似乎不一定有罪，

and more like a systemic error,

而較像體制上的錯誤，

or bias in the lineup.

或對列隊中的人有偏見。

And systemic errors don't just appear in matters of human judgement.

體制錯誤不只發生在 人為判斷的事而已。

From 1993-2008,

從 1993 - 2008，

the same female DNA was found in multiple crime scenes around Europe,

在歐洲各地許多犯罪現場 發現同一位女性的 DNA，

incriminating an elusive killer dubbed the Phantom of Heilbronn.

這位涉罪行蹤飄忽的殺手 被命名 「海布隆魅影」。

But the DNA evidence was so consistent precisely because it was wrong.

但 DNA 証據如此一致性 正因為它是錯的，

It turned out that the cotton swabs used to collect the DNA samples

原來是用來收集 DNA 檢體的棉棒，

had all been accidentally contaminated by a woman working in the swab factory.

全被一位在棉棒工廠工作的女性 不經意地給污染了。

In other cases, systematic errors arise through deliberate fraud,

在其他情況， 體制錯誤是因於蓄意欺詐，

like the presidential referendum held by Saddam Hussein in 2002,

例如 2002 年，薩達姆•海珊主導的 全民公投總統大選，

which claimed a turnout of 100% of voters with all 100% supposedly voting in favor

宣稱投票率 100%， 且據稱有 100% 是贊成

of another seven-year term.

他另一個七年任期。

When you look at it this way,

從這個方面考量時，

the paradox of unanimity isn't actually all that paradoxical.

「一致性悖論」事實上 並不是全然荒謬的。

Unanimous agreement is still theoretically ideal,

一致同意仍是理論上的理想，

especially in cases when you'd expect very low odds of variability and uncertainty,

尤其在一些情況 預期變異及不確定的機率很低時，

but in practice,

但實際上，

achieving it in situations where perfect agreement is highly unlikely

在極不可能完全一致的情況下 卻出現一致時，

should tell us that there's probably some hidden factor affecting the system.

這告訴了我們 可能有些隱藏因素影響了體制。

Although we may strive for harmony and consensus,

雖然我們力求和諧與共識，

in many situations, error and disagreement should be naturally expected.

但許多情況， 錯誤與分歧理當是預料中之事，

And if a perfect result seems too good to be true,

若一個完美結果好得令人難以置信， 它大概真的其中有詐了。

it probably is.

翻譯：Helen Lin