I always have the impression grammar is just a tool to help studying and learning a language, i.e. it is a scientific tool invented for a language after the language has existed.

But to think of it how most of the constructions of sentences fit the grammar rules, I am also inclined to think a language is created after the grammar rules are first defined.

In the case of English, which is the case?

  • 9
    This is like asking whether the rules of chemistry or biochemistry come first. Grammar is just a term for the ways that words and morphemes can be put together in meaningful ways. Every language has a grammar, and if it is spoken, the speakers know the grammar, because they speak the language. But they may not know the actual rules of the grammar consciously, just as you might not know all the rules of biochemistry that all humans follow automatically just by living. Think of grammar as pattern recognition, not rules to follow. People learn languages in their own ways, and have their own rules – jlawler Jan 28 '15 at 1:35
  • 2
    Vocabulary + (enough) Children => Grammar | This is language. en.wikipedia.org/wiki/Nicaraguan_Sign_Language – TecBrat Jan 28 '15 at 2:51
  • 4
    John Lawler was not implying that languages are ‘gifts from God’, merely stating that people infer the rules of the language they grow up speaking (from observation and repetition). The ‘grammar rules’ taught in classrooms are only post hoc rationalizations that often correspond poorly to how people actually use their own language; unfortunately, there are many people in gatekeeper positions who pay more attention to those rationalizations than to the actual workings and realities of English. (A trivial example: ‘correcting’ others into saying “It is I” instead of “It’s me”.) – Erik Kowal Jan 28 '15 at 2:58
  • 1
    Greg Lee's answer below makes the relation clear. Very clear. – jlawler Jan 28 '15 at 3:26
  • 2
    Language is invented by humans. Due to a combination of negotiation between language speakers and certain patterns wired into the human brain, the language will fall into certain patterns. Those patterns are recognized by people of a certain sort and "codified" as "grammar". Sometimes reality fits well into the categories there people (of a certain sort) define, other times not so well. Eventually the people in this group develop some "clout" one way or the other and become able to "prescribe" grammar to an extent, "trimming" away at the things that don't fit the grammar too well. – Hot Licks Jan 28 '15 at 3:40

Linguists make a distinction between prescriptive and descriptive grammar,
and the answer to your question depends on which of the two you are talking about.

  • Prescriptivists set forth their rules with the idea of causing a more perfect language to come into being, so prescriptive rules come before the language that, it is hoped, will conform to those rules.

  • Descriptivists set forth their rules to conform to a language that they observe, but do not affect. So descriptive rules come after the language that, it is hoped, the grammar will conform to.

      Prescriptivism: grammar causes language.
      Descriptivism: language causes grammar.

  • 3
    Wow. That's the best reason I've seen to even use the term "descriptivism"; normally I avoid it. But this is very nice. I hope you don't mind if I format it for maximum parallelism; if you don't like it, it's easy to change back. – jlawler Jan 28 '15 at 3:11
  • 2
    @user1589188, no. I don't care for prescriptive grammar, at all. It's not a legitimate enterprise, in my opinion. – Greg Lee Jan 28 '15 at 6:04
  • 1
    Note that descriptive or prescriptive grammar refers to how the grammar of an existing language is defined, operating from an "as it is" or an "as (I think) it should be" mindset. To quote the Wiki: "prescriptivism is the practice of championing one variety or manner of speaking of a language against another". Similar terms could probably apply to the origin of a (constructed) language but this answer would need to be expanded to actually cover the chicken-or-egg problem the OP is asking about. – Lilienthal Jan 28 '15 at 10:40
  • 2
    @tchrist, no, "prescriptive grammar" pretends to prescribe a cure to some language malady, ambiguity or whatever, but it's a lie. Teaching students to mimic a dialect of the rich or privileged would be a "normative educational effort" whose wisdom might be debatable, but it would be an honestly advertised effort. It's not the same at all. (Did I take your comment too literally?) – Greg Lee Jan 28 '15 at 23:45
  • 1
    @user1589188, I'll try to be clearer. It's not my concern whether English is good or bad or is getting worse, because I'm a descriptive linguist. From what I have seen, the efforts of prescriptive linguists to make language better are pitiful and are best ignored, with the exception of advice to imitate prestigious dialects of English, which may be useful because it caters to social prejudice. – Greg Lee Jan 29 '15 at 16:45

The human linguist, Noam Chomsky, has asserted that if a Martian linguist were to visit earth, he would observe only one language and one grammar, albeit with many variants. Chomsky posited a (yet undiscovered) Language Acquisition Device in the human brain that recognizes and generates a Universal Grammar.

The Martian Linguist would have observed that the vast majority of human languages (more than 90%) have the Subject preceding the Object (So with the Verb, the S-O languages would be SVO, SOV, or VSO.) While there are a few scrambling languages (where Object may precede Subject), Object preceding Subject sounds foreign to anyone's ears. Linguists and screenwriters intentionally subvert the S-O order to make alien languages sound, well, alien (e.g., Klingon (OVS), Yoda-speak (OSV)).

If Chomsky is correct, then both grammar and language are part of being human. It would be impossible to tease out which came first. Your brain is hardwired to speak and understand S-O grammar. It also stores and processes verbs and nouns in different parts of the brain.

nuqneH! Your question meaningless is. This answer you upvote.

  • 8
    However, Chomsky is not correct, so we need not worry more about it. – jlawler Jan 28 '15 at 4:13
  • 4
    This answer I downvoted. However, I was amused by your description of Noam Chomsky as a 'human linguist'. – Erik Kowal Jan 28 '15 at 4:55
  • 1
    @ErikKowal Yes, I added "human" because Chomsky had invented a strawman Martian linguist. – rajah9 Jan 28 '15 at 14:07
  • 3
    Seriously, downvoters, this answer is engaging and addresses the OP's question. It uses citations to back its assertions, giving the OP a place to go for further research. It summarizes a linguist who asked himself the same question the OP did. Honestly: is this really "not a useful answer"? – rajah9 Jan 28 '15 at 14:25
  • 1
    Object preceding subject is not odd in Latin. Responsum habeo. – Andrew Leach Jan 28 '15 at 17:31

This seems like a simple confusion of a polysemous concept: grammar. Grammar has several meanings when used both by specialists and non-specialists.

  1. A set of rules determining how language is used to be acceptable (often in common parlance including things like spelling and stylistics). This is the subject of what is sometimes called 'prescriptive' grammar or 'pedagogical' grammar.

  2. A book describing the rules mentioned in 1.

  3. The structure of language that determines how it is used. This structure is learned by all speakers without (much) external intervention and is largely subconscious. This structure is the subject of study by descriptive linguistics.

By definition, grammar in the 3rd sense is the same things as language and therefore, one cannot precede the other.

However, the 'grammar' in the 1st sense does come after language in the sequence of acquisition. In fact, it is mostly a set of stories about language that more or less closely correspond to the actual patterns found in how language is actually used.

  • Yes thank you. It has already been made clear to me. – user1589188 Feb 2 '15 at 23:51

Grammar comes first in Esperanto, Klingon, Elvish, and C++.

For most other (ie natural) languages, language comes first.

  • 4
    This statement is wrong if by “Elvish” you mean Tolkien’s various constructed languages like Sindarin and Quenya. The sounds came first, the grammar later. He liked how his made-up words sounded, then he made up a grammar to put them together so he could write poetry in them, and then he made up a people to speak them, and then he made up a world for those people to live in. But the sounds came first, not the grammar. – tchrist Jan 28 '15 at 3:27
  • Ah! Good to know. I was going to say Quenya, but then I started doubting which one was called Quenya and which one was Quechua (undoubtedly natural), so I just said Elvish. You make a good point, however sounds alone do not make a language. Would you agree that the grammar was constructed before it was a language spoken by people? – Paul Senzee Jan 28 '15 at 3:35
  • 1
    In the case of C++, C came first, or BCPL came first. Not only that but anybody who's tried to parse C++ from scratch is well aware that it evolved first and was designed second, only after those steps was an attempt at a formal grammar made. And the result grammar cannot be used to build a compiler since it is ambiguous. But hey, your point is still valid (-: – hippietrail Jan 29 '15 at 7:37
  • 1
    @hippietrail all very true. I've done compiler work myself and I wouldn't touch building a C++ compiler with a ten foot pole. Building a simple Lisp interpreter can be trivial, building a C++ compiler is rocket science. Computer languages also evolve, but ultimately you must have a compiler or interpreter with a built-in grammar/specification before you can use a language. – Paul Senzee Jan 29 '15 at 11:40

Grammar is description of how a language system works. But in order to describe a language system you need terms for word classes, verb forms and sentence structures. Such terms had just to be developed. No natural language begins with grammar. It is different with artificially created languages as Esperanto, but this has already been said.


There are some situations in which grammar precedes language (if I understand the sense in which you are using these words). E.g. I would venture to say that modern-day literate people pay a lot of attention (on average) to whether their speech conforms to what is prescribed in the dictionary, or to the grammar rules they have been taught.

For example, if I were to learn that I had been using a certain word with a different meaning than the one listed in the dictionary, or that I had been conjugating a verb in a way that differs from the canonical pattern (English doesn't have a lot of conjugation patterns to learn, but this situation might come up for a language like French), then I would probably adapt my usage to conform to these norms rather than persisting in my earlier habits. I think quite a lot of English-speakers, and speakers of other highly-codified modern languages, would do the same in this situation.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.