到底什么算一个词

我刚才在手机上看到一句话，一开始怎么看都看不懂，后来才发现，是我把里面的词语切错了地方，换一种切分方法，意思一下就通了。

I just saw a sentence on my phone that I couldn't make sense of at first. Later I realized I had split the words in it apart in the wrong place. Once I segmented it a different way, the meaning suddenly clicked.

这种事我也常常遇到。中文写出来，字和字中间没有空格，一个词语从哪个字开始、到哪个字结束，很多时候得靠自己判断，这其实是语言学里一个很有意思的问题。你说的那句话是什么？

That happens to me a lot too. When Chinese is written out, there's no space between one character and the next, so where a word actually starts and where it ends often has to be judged by the reader. That's actually a pretty interesting question in linguistics. What was the sentence you saw?

是"研究生命起源"这几个字。我第一次看，脑子里自动切成"研究生"和"命起源"，可是"命起源"完全说不通，我想了半天才明白，应该切成"研究"和"生命起源"，是在说研究生命的起源，这样才对。

It was the characters 研究生命起源. The first time I looked at it, my mind automatically split it into 研究生 and 命起源, but 命起源 doesn't make any sense at all. I thought about it for a while before realizing it should be split as 研究 and 生命起源, meaning to study the origin of life. That's the reading that actually works.

这个例子我知道，是语言学里常常用来讲切分问题的一句话。同样几个字，一种切分是"研究生"加"命起源"，读不通；另一种切分是"研究"加"生命起源"，才是一句正常的话。这就是切分带来的歧义，同样的字，切的地方不一样，意思完全不一样。

I know this example. It's a sentence linguistics often uses to talk about the segmentation problem. With the same characters, one split gives you 研究生 plus 命起源, which doesn't read as sense; the other split gives you 研究 plus 生命起源, which is a normal sentence. This is exactly the ambiguity segmentation creates. Same characters, different cutting point, completely different meaning.

这么说来，是不是只要脑子聪明一点，就能一次把词语切对，不会切错？

So does that mean if your mind is just a bit sharper, you'll segment the words right the first time and never get it wrong?

不是聪明不聪明的问题，这句话本身就有两种切分方法，两个都合语法，只是一个说得通意思，一个说不通。这种歧义也不是只有这一句话才有，很多句子只看字面，两种切分都讲得通，谁也没办法一下断定哪种是对的，就算你的母语是中文也一样。

It's not about being smart or not. This sentence genuinely has two possible segmentations, both grammatically valid. It's just that one makes sense and the other doesn't. And this kind of ambiguity isn't limited to this one sentence either. In a lot of sentences, if you just look at the characters on the page, both segmentations work, and there's no way to immediately settle which one is correct, even if Chinese is your native language.

那有没有真的说不清哪种对的句子？不是像"研究生命起源"这样一种说得通、一种说不通，是两种切分都说得通，可是意思完全不一样的那种歧义？

Is there a sentence where you really can't tell which way is right, not like 研究生命起源 where one makes sense and the other doesn't, but the kind of ambiguity where both segmentations make sense, yet mean completely different things?

有，比如"乒乓球拍卖完了"这几个字。一种切分是"乒乓球拍"加"卖完了"，意思是打乒乓球用的球拍卖完了；另一种切分是"乒乓球"加"拍卖"加"完了"，意思是乒乓球的拍卖已经结束了。两种切分都说得通，语法上也都没问题，可是说的完全是两件事。

Yes, take the characters 乒乓球拍卖完了. One segmentation gives you 乒乓球拍 plus 卖完了, meaning the paddles used for playing ping pong are sold out. The other gives you 乒乓球 plus 拍卖 plus 完了, meaning the auction of ping pong balls has ended. Both segmentations make sense, and neither has a grammar problem, but they're talking about two completely different things.

这个例子太有意思了，一个是体育用品卖完了，一个是拍卖会结束了，两件事一点关系都没有，可是写出来的字一模一样。这样看，中文句子写出来以后，到底想表达哪个意思，有时候真的没办法只靠字面判断。

This example is fascinating. One is sporting goods being sold out, the other is an auction ending, and the two things have nothing to do with each other at all, yet the characters written down are exactly the same. Looking at it this way, once a Chinese sentence is written down, sometimes you really can't judge which meaning is intended just from the characters on the page.

对，而且这不是只有考试题目里才有的怪句子。我以前还看过一个更有名的例子，是"南京市长江大桥"这几个字。一种切分是"南京市"加"长江大桥"，说的是南京市的长江大桥；另一种切分是"南京市长"加"江大桥"，好像是在说一个姓江的市长。两种切分读出来完全不一样，可是字是同一串字。

Right, and this isn't only the kind of odd sentence you find in test questions. I've also seen an even more famous example before, the characters 南京市长江大桥. One segmentation gives you 南京市 plus 长江大桥, referring to Nanjing city's Yangtze River bridge. The other gives you 南京市长 plus 江大桥, which sounds like it's talking about a mayor surnamed Jiang. The two segmentations read out completely differently, yet it's the same string of characters.

那到底应该听谁的？有没有一个标准答案，规定中文句子必须这样切分，不能那样切分？

So who should we go by in the end? Is there a standard answer that dictates a Chinese sentence must be segmented this way and not that way?

这正是语言学里到现在还在讨论的问题。中文写出来，字和字之间本来就没有空格，一个词语从哪里开始、到哪里结束，很多时候不是写出来就定死了的，而是每个人一边读一边自己切分，切的方法还可能不一样。母语是中文的人，看到同一句话，有时候切分的方法都会不一样，更不用说想让电脑自动切分中文句子有多难了。

This is exactly the kind of question linguistics is still discussing even now. When Chinese is written down, there's simply no space between one character and the next to begin with. Where a word starts and where it ends is, in a lot of cases, not fixed once and for all by the writing. Instead, each person segments it themselves while reading, and the way they segment it can differ. Even native Chinese speakers, looking at the very same sentence, sometimes segment it differently from each other, not to mention how hard it is to get a computer to automatically segment a Chinese sentence.

你刚才说，母语是中文的人也可能切错。这么听你说完，我以后看一句话看不懂的时候，说不定不是我中文不好，是这句话本来就有歧义，能切成两种样子。

You just said even people whose native language is Chinese can segment it wrong. Now that I've heard you explain all this, next time I read a sentence and can't make sense of it, maybe it's not that my Chinese isn't good enough. Maybe the sentence itself genuinely has an ambiguity and can be split two different ways.

说得对，先别急着怀疑自己，先想想是不是还有另一种切分的可能。

Exactly right. Don't rush to doubt yourself first. First think about whether there's another possible way to segment it.