如果我们遇到外星人，我们能理解他们吗？

    导读：让我们想象一下，如果外星人突然出现在我们面前，我们首先会做什么？当然，优先要做的事情肯定是和平交流。但是我们能听懂彼此的话吗？

    许多科学家相信外星文明的存在。对他们来说，现在的问题是，我们是否会在不久的将来或距今一段时间之后遇到他们，而不是如果。因此，让我们想象一下，如果外星人突然出现在我们面前，我们首先会做什么？当然，优先要做的事情肯定是和平交流。但是我们能听懂彼此的话吗？

我们肯定会和外星人交流科学信息。如果在整个宇宙中，宇宙的定律都是相同的，那么对于这些定律，外星人应该会有和我们不一样的表述方式，但是原则上，定律都应该是相同的。

    提到语言，问题就更复杂了，语言是人类合作中最最重要的因素。只有通过交流沟通，我们才能明白彼此的意图，这是一个大的团队进行合作的基础。正因为如此，任何技术上先进的外星文明都应该拥有类似语言的能力。

那我们有可能去学一门外星语吗？首要的障碍是媒介。人类是在85-255赫兹的音频范围内和430-770太赫兹的光频范围内交流。对外星人来说，可能不是这样，他们有可能与我们具有不同的进化路径。这很大程度上是个技术问题。举个例子，将鲸鱼的声音加快，人类就听不到它们的声音，也就是说，把外星人的声音映射成某种人类无法察觉的形式，很容易做到。

    语法和语义

    更为棘手的问题是，我们是否能学会一种外星人语言的内部结构。现有的语言心理学研究给出了两个非常不同的答案。

生成语法学家认为，语言结构是根植于大脑的，这表明学习外星语是不可能的。该观点认为，人类天生就具有一种普遍语法，它实质上是一种大脑具有的与语言知识相关的特定状态，在一个给定的语言基础体系中生成各种具体语言。我们童年听到的语言激活这些状态（也就是后天语言环境不同），使我们能够区分有效和无效的组合词的方式（在普遍语法范围内获得各不相同的母语生成语法）。

关键的一点是，普遍语法不是语法大全。虽然人类的普通语言包罗万象，但是普遍语法是一些严格的条件，能够限制人类语法的可能范围。例如，“动词方向性”条件决定一种语言中的动词在补语之前或之后，英语是动词在前面（鲍伯将一块蛋糕给了爱丽丝）而日语是将动词放在最后，（按照鲍伯 爱丽丝 蛋糕 给了这个顺序）。

    对于生成语法学家来说，外星人碰巧与人类拥有相同的条件，这是极不可能的。用这种观点的主要倡导者，诺姆·乔姆斯基的话来说：

如果来自外太空的火星人登陆地球，说着违背我们普遍语法的语言，我们根本就无法按照学习人类语言像英语和斯瓦希里语这样的方式去学习它…我们天生就能学习英语，汉语，和其他所有的人类语言。但我们不能学习违反普遍语法的语言。

另一方面，认知派观点认为语义比语法更重要。根据这一观点，这样的句子：“四倍 喝酒 拖延quadruplicity drinks procrastination ”合乎语法但是不具有语义意义。因为这个原因，认知派观点的支持者认为，了解一门语言仅靠语法是不够的。相反，它需要辅以一定的知识，就是该语言使用者思考的结构。

    我们也可以看看周围的世界，为什么生物有惊人的相似之处，即使它们在不同环境中以不同的方式进化。这被称为“趋同进化”（间隔十万八千里，基因不同的物种，为了适应相似的外部环境，越来越趋于同质化了）。例如，自然界术语中，在动物界，翅膀和眼睛都是通过不同时期的进化独立出现的，而且在生态隔离的新西兰，鸟类进化行为在其他地方的哺乳动物身上也能看到。认知派的观点为人类和外星人的语言相互理解提供了希望。

但认知派的观点正确吗？神经网络的研究表明，语言可以被学习，大脑中没有专门的结构。这点很重要的，因为它意味着可能没有必要假设一个先天的普遍语法来解释语言习得。此外，有可能是人类的语言不符合普遍语法框架。虽然这些结果还远远没有定论（例如，他们不能解释为什么只有人类有语言），证据倾向于认知派的观点。

    因此，假设人类可以学习外星语，这个假设具有合理性。显然，可能外星语中的一个方面（像我们的诗歌一样）是我们永远无法理解的。同样，一些物种可能也拥有一个不同的精神世界，与人类的大致相当。然而，我认为，我们可以谨慎乐观地认为，在物理、生物和社会世界里，总有一个方面能够让人类和外星语被框定在一个共同的语义框架内。

    “英文原文”

    If we ever came across aliens, would we be able to understand them?

    Many scientists believe that alien civilisations exist. For them, the question is now whether we will encounter them in the near future or a very long time from now, rather than if at all. So let's imagine that we suddenly stand face-to-face with members of an alien species. What would we do first? Surely communicating that we come in peace would be a priority. But would we ever be able to understand each other?

    The one thing we can be confident about exchanging with aliens is scientific information. If the laws of the universe are the same everywhere, then different descriptions of these laws should, in principle, be equivalent.

    Matters are more complicated when it comes to language, which is the single most important factor in human cooperation. It is by communicating our intentions that we are able to work together in surprisingly large groups. For this reason, it is plausible that any technologically versatile alien civilisation would have something like language.

    Can we expect to learn such an alien language? The first hurdle would be its medium. Humans communicate in a 85-255Hz frequency range of sound and in the 430-770 THz frequency range of light. This is unlikely to be true of aliens, who will have evolved differently. Nevertheless, the problem is largely a technical one. Speeded up whale songs that are otherwise inaudible to humans, for instance, show that it is relatively easy to map "alien" stimuli into forms that humans can perceive.

    Grammar versus semantics

    The more difficult question is whether we would ever be able to learn the internal structure of an alien language. Existing perspectives in the psychology of language give two very different answers.

    The generativist approach, which holds that the structure of language is hardwired into the brain, suggests this wouldn't be possible. It argues that humans come with an inbuilt universal grammar that has a specific number of settings – each corresponding to the acceptable order in which words and parts of words can be arranged in a given language system. The language we hear in early life activates one of these settings, which then allows us to distinguish between valid and invalid ways of combining words.

    The key point is that the number of grammars is very limited. Though the rules of human languages can and do vary, proponents of the generativist model argue they can only do so within strict parameters. For example, the "head directionality" parameter determines whether the verbs in a language precede or follow their complements, with English being head-initial ("Bob gave a cake to Alice") and Japanese being head-final ("Bob to Alice a cake gave").

    For generativists, it is extremely unlikely that an alien species would happen to have the same parameters as human beings. In the words of Noam Chomsky, the leading proponent of this view:

If a Martian landed from outer space and spoke a language that violated universal grammar, we simply would not be able to learn that language the way that we learn a human language like English or Swahili … We're designed by nature for English, Chinese, and every other possible human language. But we're not designed to learn perfectly usable languages that violate universal grammar.

    The?cognitive view, on the other hand, sees semantics (structures of meaning) as being more important than syntax (structures of grammar). According to this view, sentences like "quadruplicity drinks procrastination" are syntactically well-formed but semantically meaningless. For this reason, proponents of the cognitive view argue that grammar alone is not enough to understand language. Instead, it needs to be partnered with knowledge of the concepts that structure how language users think.

    We can also look at our own world to see how organisms can have striking similarities, even if they have developed in very different ways and in contrasting environments. This is known as "convergent evolution". In physical terms, for example, wings and eyes have independently emerged among animals through evolution at several different times over, and birds in ecologically isolated New Zealand have evolved behaviours typically seen in mammals elsewhere.The cognitive view offers hope that human and alien languages might be mutually intelligible.

    But is the cognitive view correct? Research on neural networks shows that languages could be learned without specialised structures in the brain. This is important because it means there may be no need to postulate an innate universal grammar to explain language acquisition. Also, it seems there may be human languages that don't fit in the universal grammar framework. Though these results are far from conclusive (for instance, they can't explain why humans alone seem to have language), the evidence leans towards the cognitive account.

   So, it might be reasonable to assume that humans could learn alien languages. Clearly, there would probably always be aspects of an alien language (like our poetry) that are inaccessible. Equally, some species may occupy such a different mental universe that it is only broadly equivalent to that of humans. Nevertheless, I think we can be cautiously optimistic that universal structures in the physical, biological and social worlds would be enough to anchor human and?alien?languages in a common semantic framework.