故事从一本词典开始 / It all starts with a dictionary
说实话，我第一次知道 Big Five 是怎么"被发现"的时候，有点震惊。我以为是哪个心理学家闭关三年想出来的理论模型，结果——它的起点是一本英语词典。
Honestly, when I first learned how the Big Five was discovered, I was a little stunned. I'd assumed some psychologist sat in a cabin for three years and dreamed up a model. Nope. The starting point was a dictionary.
1936 年，Allport 和 Odbert 从英文词典里翻出了大概 4500 个和性格相关的形容词——勤奋的、好奇的、敏感的、爱炫耀的、爱抱怨的、磨叽的——然后开始做一件极其朴素的事：把这些词反复压缩、合并、做因素分析，看看到底能合并成几个独立的维度。这个思路叫"词汇假设"（lexical hypothesis）：人类最重要的性格差异，应该早就被我们的日常语言编码进去了。如果某个特质真的重要，就一定有人发明词去描述它。
In 1936, Allport and Odbert pulled roughly 4,500 personality-related adjectives out of the English dictionary — diligent, curious, sensitive, show-offy, whiny, slow-moving — and then did something almost embarrassingly basic: they kept compressing and merging the list and running factor analyses, to see how many genuinely distinct dimensions the words boiled down to. The underlying idea is called the lexical hypothesis: if a personality trait really matters, language will already have encoded it. Important things get words.
几十年后，经过 Cattell、Tupes、Norman、Goldberg、McCrae、Costa 这一串名字的反复挖掘，结果稳定地掉出了 5 个大因子：开放性、尽责性、外向性、宜人性、神经质。OCEAN。Big Five 就这么诞生了。
A few decades and a long lineage of researchers later — Cattell, Tupes, Norman, Goldberg, McCrae, Costa — the same five factors kept falling out: Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism. OCEAN. That's how the Big Five was born.
听上去挺漂亮的对吧？但你有没有注意到一个小问题：整个故事的起点，是英文词典。
Sounds elegant, right? But you might have noticed a small problem: the whole story starts with an English dictionary.
"词汇假设"的语言依赖性 / The lexical hypothesis is language-bound
如果"重要的性格特质都会进入语言"这个假设是对的，那不同语言应该独立得出相似的五维结构。这是个可以验证的预言——心理学家们也确实做了几十年这事。
If the lexical hypothesis really holds, then different languages should independently produce a similar five-factor structure. That's a falsifiable prediction, and psychologists have spent decades testing it.
好消息是，在德语、荷兰语、意大利语、波兰语、捷克语、菲律宾语、土耳其语这些语种的研究里，Big Five 的结构基本可以复制——尤其是开放性、尽责性、外向性、神经质，跨语言稳定性相当高。听起来是个胜利吧？
The good news: in German, Dutch, Italian, Polish, Czech, Filipino, Turkish, and many others, the Big Five structure largely replicates — Openness, Conscientiousness, Extraversion, and Neuroticism in particular are remarkably stable across languages. Looks like a win, doesn't it?
但仔细看，会发现两件事：第一，能"基本复制"的那些语种，绝大多数属于印欧语系；第二，宜人性（Agreeableness）这个维度在很多文化里都不那么干净——它有时候会拆开、有时候和别的维度纠缠在一起。这就开始有点意思了。
But look closer and two things stand out. First, the languages where it "basically replicates" are almost all Indo-European. Second, Agreeableness in particular tends to get messy across cultures — it sometimes splits, sometimes blurs into adjacent factors. Things start to get interesting.
WEIRD 样本的尴尬 / The WEIRD-samples problem
2010 年有篇影响力极大的文章，作者是 Henrich、Heine、Norenzayan，标题叫《The Weirdest People in the World》。他们指出心理学这门学科的一个尴尬事实：它的研究样本绝大多数来自 WEIRD 群体——Western（西方）、Educated（受过教育）、Industrialized（工业化的）、Rich（富裕的）、Democratic（民主的）。然后把这些样本得出的规律，当成"人类心理学"四个字普遍适用。
In 2010, Henrich, Heine, and Norenzayan published a paper that landed like a small bomb in psychology: The Weirdest People in the World. Their point was uncomfortable: most psychology research uses samples from a very narrow slice of humanity — WEIRD: Western, Educated, Industrialized, Rich, Democratic. And then we generalize those findings as if they describe Homo sapiens in general.
Big Five 也吃了这一刀。它的早期数据，大多数来自美国和西欧的大学生——很难想象比这更 WEIRD 的样本了。所以当研究者把它扛到中国农村、非洲部落、拉美工人阶级社区去验证，结构开始变形也就不那么意外了。
The Big Five caught a piece of this. Its early data came overwhelmingly from American and Western European university students — it's hard to imagine a more WEIRD sample. So when researchers took the model to rural China, African tribal communities, or working-class Latin American neighborhoods, it's not really shocking that the structure started to bend.
CPAI：中国人发现的"第六维度" / CPAI: the sixth dimension Chinese researchers found
这是这篇文章里我最想讲的一段。上世纪 90 年代，香港中文大学的张妙清（Fanny Cheung）等人不满意 Big Five 直接套到中国语境的效果，决定从零开始——用中文做整个词汇分析，看会得出什么结构。
This is the part I most wanted to write about. In the 1990s, Fanny Cheung and her colleagues at the Chinese University of Hong Kong weren't satisfied with just translating the Big Five into Chinese. So they started from scratch — running the entire lexical analysis in Chinese to see what would fall out.
他们造出来的东西叫 CPAI（Chinese Personality Assessment Inventory，中国人个性测量表）。除了能对应上 Big Five 的大部分维度，他们还稳定地发现了一个独立的因子——Interpersonal Relatedness（人际关系性）。它包含的内容大概有：和谐（不破坏关系的能力）、人情、面子、关系灵活性、传统主义……
What they built is called CPAI — the Chinese Personality Assessment Inventory. Most of its factors line up with the Big Five, but they consistently extracted one extra independent factor: Interpersonal Relatedness. The cluster includes things like harmony (avoiding rupture in relationships), renqing (the give-and-take of social favors), face, relational flexibility, traditionalism…
听起来很"东亚"对吧？但他们后来在西方样本上跑 CPAI 的时候发现：这个因子在西方样本里也存在，只是西方的人格研究者从来没把它当成一个独立维度去挖。换句话说，不是西方人没有"人情面子"，而是英语词典里这部分词汇没多到能被因子分析单独抓出来。
Sounds very East Asian, right? But here's the twist: when they ran the CPAI on Western samples, the factor showed up there too — it just had never been pulled out as a separate dimension by Western researchers. In other words, Westerners aren't immune to renqing and face dynamics; it's just that the English-language vocabulary for it isn't dense enough to fall out of an English factor analysis.
（这一段我读到的时候我其实有点感慨。所谓"客观的"科学发现，背后有多少是被工具——比如你用哪种语言、哪本词典——给框住的。）
(I'll admit, I had a moment when I first read this. So much of what we call "objective" scientific finding turns out to be quietly shaped by the tools — which language, which dictionary, which sample.)
那是不是说 Big Five 错了？/ So is the Big Five wrong?
不是。这事吧，不能用对错来评。
No. The right framing isn't "right or wrong."
更准确的说法是：Big Five 抓到的是人类性格差异里相当大的一块，但不是全部。它在跨文化里能稳定复制 4 个维度（OCEN）、宜人性次之、还漏掉了一些和"集体"、"关系"、"宗教性"相关的维度。它是个很好但不完整的地图。
A better framing: the Big Five captures a large chunk of human personality variation, but not all of it. It cross-culturally replicates four of its five dimensions pretty cleanly (OCEN), Agreeableness more shakily, and it misses some dimensions tied to collectivism, relational identity, and religiosity. It's a good but incomplete map.
这也是为什么后来出现了 HEXACO ——它加了一个"诚实-谦逊"维度，部分弥补了 Big Five 对道德相关特质的低估。同样的逻辑，CPAI 加的是"人际关系性"。模型在演化。
This is one reason HEXACO emerged — it adds an Honesty-Humility factor that the Big Five undersamples around moral traits. The CPAI does something similar by adding Interpersonal Relatedness. The models are evolving.
对你做测试的意义 / What this means for the tests you take
讲真，如果你做的是国内一些"伪 Big Five"短测——比如某些公众号上 5 分钟测出来的"OCEAN"——它的题目是中文，结构却照搬英文版本。这种测，很可能在"人际关系性"这部分对你完全无感。它没有这些题，自然测不到这个维度上。
So here's the practical takeaway. If you've taken one of those quick 5-minute "OCEAN" quizzes floating around Chinese WeChat — most of them are translated directly from the English version. They likely miss the Interpersonal Relatedness dimension entirely. Their items don't ask about it, so they can't see it.
我们做 SBTI 的时候，故意没有走严格的 Big Five 路线——一方面是因为我们不是临床工具，另一方面也是因为我希望保留一些"东亚特有"的观察角度（比如"自我合理化能力"、"社交雷达灵敏度"）。这些不是经过严格因子分析的科学结构，但它们恰好捕捉到了 Big Five 在中文语境里漏掉的东西。
When we designed SBTI, we deliberately didn't take the strict Big Five route — partly because we're not a clinical tool, and partly because I wanted to keep some "East Asian-flavored" observation angles (like "self-rationalization capacity," "social radar sensitivity"). These aren't validated psychometric constructs. But they happen to capture some of what the Big Five misses in a Chinese-speaking context.
如果你真的想严肃测一次 Big Five，去找 IPIP-NEO 或者 BFI-2 的中文版——它们至少经过了一些本地化验证。但要记住：再严肃的工具也只是一个切片，不是你本人。
If you really want a serious Big Five assessment, look for the Chinese-validated versions of IPIP-NEO or BFI-2 — they've at least gone through some local adaptation. But remember: even the most rigorous instrument is a slice, not you.
一个不太收尾的收尾 / An open-ended ending
这篇我本来想给一个干脆的结论，但写到最后越想越觉得，跨文化人格研究最有意思的地方就是——它一直在提醒我们，所谓"普世的人类心理"，其实带着一层很厚的文化滤镜。我们今天读到的所有人格理论，都是某个具体语言、某个具体年代、某个具体阶级写出来的。这不是要否定它们，而是要带着一点警觉去读。
I wanted to wrap this up with a clean conclusion. The more I wrote, the harder that got. What's most interesting about cross-cultural personality work is that it keeps reminding us: "universal human psychology" is filtered through a pretty thick cultural lens. Every personality theory you've read was written in a specific language, in a specific era, by a specific class of people. That doesn't mean reject them. It means read them with a little caution.
下次再有人跟你说"研究显示外向的人更幸福"——你可以心里多问一句：在谁的研究里？哪国的样本？什么年代？
Next time someone tells you "research shows extroverts are happier" — feel free to mentally append: whose research? what country's sample? what decade?
本文是科普与个人观察材料，不构成专业建议。This piece is for educational and reflective purposes; it is not professional advice.