Disclaimers: I have no linguistic knowledge whatsoever, I'm just fascinated by these subjects. Also, I will use the word "dialect" due to my lack of a better word, although I see that the description of the "dialects" tag talks about "mutually intelligible" while the things I'm talking about don't need to be mutually intelligible and in fact are often not. I hope someone can correct my use of the word "dialect" with the right word, but this is not my question, just a premise.

The context I have in mind is Italy, where we have a national language taught in school and a myriad of very localized dialects, which are full-blown languages with their syntax, pronunciation, rules etc. and which are often almost unintelligible by speakers of other dialects and by Italian-only speakers. But I'm sure the same phenomenon exists in many countries so my question is not specific to Italy at all.

Up until a few decades ago, most of these dialect speakers, especially the older persons and especially in the countryside, barely went to elementary school.

Most of these dialect speakers make a significant amount of errors when they speak Italian, but they speak their dialect basically without any errors at all. In the same geographical area these speakers all strictly comply with the same language rules and their compliance is very consistent among the population and the level of this compliance is certainly not inferior (and maybe superior) to the level of compliance of the average urban Italian high-school graduate speaking Italian.

What puzzles me is that their dialect was not taught to them in school (where most likely it was rather discouraged and mocked) and I doubt that people in their community taught them the rules of the dialect. I believe that the children just learn it by being raised where it is commonly spoken.

This clashes with my long-held idea that in order to learn to speak a language - even one's native language - with perfect compliance with the rules, one must have had the rules taught to him/her, while this doesn't happen for these dialect speakers, they don't have the rules taught by others.

I would expect that if one learns a language just by growing up where it's spoken, without being taught any rules, of course one would learn it (especially if one starts as a child) but this would result in a general sloppiness regarding compliance with the rules among the speakers, or even result in a scarcity of actual rules. Instead no: precise rules, full compliance, hardly an error.

So how does this aspect of languages work ? How can these dialect speakers consistently get the rules of their dialect perfectly right without having ever studied those rules ? Of course they learned that by being born and raised in a community that speaks that dialect, but why in order to get the same level of compliance with the rules in the national language we need to study it for good and still many of us hardly get as perfect at it as those dialect speakers are all perfect at their dialect ?

  • 20
    How would people have learned to speak before the rules were codified? Commented Aug 9, 2020 at 2:33
  • 10
    The assumption that you need to know the language rules to speak it seems... weird. I've always felt native speakers can't usually recite most rules of their language, but people who learned it later in life can.
    – Mark
    Commented Aug 9, 2020 at 10:15
  • 4
    @SantiBailors the "rules" of a dialect are defined by the practice of how these people speak. They are in perfect compliance by definition, any mismatch between their speech and some description of the rules simply implies that this description of rules is faulty and does not match reality.
    – Peteris
    Commented Aug 9, 2020 at 12:48
  • 17
    You have it completely backwards, we have discovered the 'rules' of languages by listening to people speak, and how they use the language, not the reverse. So it's not surprising that people speak their own language 'correctly'.
    – crobar
    Commented Aug 9, 2020 at 14:02
  • 4
    The local dialect is by definition whatever those people speak. Of course they speak whatever they speak perfectly! Commented Aug 10, 2020 at 12:28

8 Answers 8


This clashes with my long-held idea that in order to learn to speak a language - even one's native language - with perfect compliance with the rules, one must have had the rules taught to him/her, while this doesn't happen for these dialect speakers, they don't have the rules taught by others.

Indeed, children are amazingly good at learning to speak, just by being exposed to a language. Experiments like the famous "wug test" show that children figure out the rules of their native language extremely quickly, without any formal education at all.

I would expect that if one learns a language just by growing up where it's spoken, without being taught any rules, of course one would learn it (especially if one starts as a child) but this would result in a general sloppiness regarding compliance with the rules among the speakers, or even result in a scarcity of actual rules, because a perfect compliance with rules would be unnecessary and overkill for the eminently practical goals of their communications.

In fact, most of the rules of language are never taught to native speakers! One of the more famous rules of English (famous because it was the subject of the "wug test" experiment) involves the plural marker -s; sometimes it's pronounced like the S in "sip", sometimes like the Z in "zip". But this was certainly never discussed in my elementary-school English classes.

So how does this aspect of languages work ? How can these dialect speakers consistently get the rules of their dialect perfectly right without having ever studied those rules ? Of course they learned that by being born and raised in a community that speaks that dialect, but why in order to get the same level of compliance with the rules in the national language we need to study it for good and still many of us hardly get as perfect at it as those dialect speakers are all perfect at their dialect ?

There are two parts to this.

One part is learning a second language—learning standard Italian when you speak Venetian, for example. This usually involves a lot of formal education, though it's also possible to learn through a long enough period of immersion.

The other part is that a lot of rules taught in schools aren't actually part of the language (at least, not the way native speakers use it)! Writing, for example, is something that's formally taught; children learn to write much later than they learn to read.

Some other rules are prescribed by official authorities, but aren't actually used by most native speakers: English rules like "don't end sentences with prepositions" and "don't split infinitives" reflect how the authorities want people to speak, not how they actually speak. Since children grow up hearing adults say things like "what did you come up with?", they absorb the descriptive rule that prepositions can come at the end of sentences; the opposite needs to be formally taught, since it doesn't match what they've grown up with.

For many dialects/smaller languages, there's no official authority setting out prescriptive rules about spelling and grammar. Therefore, whatever native speakers grow up learning is considered correct—as opposed to languages like French, which have rules set out by the government.

  • 7
    Standard Italian doesn't have rules set out by the government, unlike some other standard languages. It's a minor point, since teachers still do teach based on textbook rules as ultimately decided arbitrarily by someone, but the Accademia della Crusca is very specifically not governmental and not prescriptive, unlike the RAE or the Académie Française (their site is actually very good with descriptive answers to many common questions).
    – LjL
    Commented Aug 8, 2020 at 19:20
  • 2
    @SantiBailors As an amateur, the fact that language speakers have internalized very specific and complex rules while being unaware of their existence sounds like decent evidence in favor of the 'actual language rules' being descriptive rather than prescriptive. But I'm not sure if my naive analysis here is on point. Commented Aug 9, 2020 at 9:06
  • 2
    @SantiBailors Indeed, it's fascinating! Do read up on the Wug Test if you haven't already, as the ur-example of young children following a rule perfectly even though that rule was never taught.
    – Draconis
    Commented Aug 9, 2020 at 16:21
  • 10
    @Discretelizard Absolutely; as a general rule of thumb, prescriptivism never works as well as people expect. English-speakers have always ended sentences with prepositions and always will. There are a few places where it's important and useful (such as in language revitalization projects), but in general, to borrow a metaphor from xkcd—prescriptivists are trying to direct a river with road signs.
    – Draconis
    Commented Aug 9, 2020 at 16:23
  • 2
    @user253751: A common phase in language development in children is that children may make mistakes that they used to never have problems with when they figured out rules for a language construct. For example, children often learnt early on to use words like "go" and "went" correctly, but then they go through a phase after they figured out they can talk about something that already happened by adding "-ed" to whatever they're doing, and go through a regression during this phase inventing constructs like "goed", which nobody around them have ever used.
    – Lie Ryan
    Commented Aug 10, 2020 at 12:54

This clashes with my long-held idea that in order to learn to speak a language - even one's native language - with perfect compliance with the rules, one must have had the rules taught to him/her, while this doesn't happen for these dialect speakers, they don't have the rules taught by others.

Good, because that's false! The human brain comes pre-wired with the ability to learn and generate human language, and it's incredibly good at doing so. There's a whole section of the brain dedicated to it. The only thing it takes to reach native level ability in any human language is a sufficient amount of comprehensible input.

Children seem really good at this because they learn their native language(s) rather quickly, and without much prompting, but it's actually adults having learnt to filter out useless information that makes children seem better. Once an adult language learner teaches their brain that their target language is important, the comprehensible input starts reaching the language centre of the brain, and it gets to work figuring it out.

The reason the people to whom you refer have a poor grasp on Italian is exactly because they don't need to speak it and listen to it all the time, and were taught it. If I say to a native English speaker, (or anyone with enough experience with the language), "the green big tree", and "the big green tree", not only will they know instantly which sentence is correct, they'll not be able to describe why (unless they happen to know that obscure grammar "rule"). They'll "just know".

To someone who learnt English from a book, as it were, they would think both sentences are equally valid, unless they'd either spent enough time listening to comprehensible input to "just know", or they'd gone so deep into the academia that they knew the grammar rule, and damningly, in the latter case, they would not be able to utilise that rule at native speed in speech. To utilise conscious language knowledge, you need enough time to think it through and remember the so-called rules. Unconscious knowledge is impossible to not utilise, however. If someone says "giraffe pimp" to me, there is no force of will I can employ to hear the sounds, and not instantly understand the meaning.

There's a little value in learning some things about a target language to aid one's journey for comprehensible input - for instance, if you come across a word or phrase being said over and over again, but don't understand it, and can't figure it out from the context, you can drop it into google images, or a native language dictionary, and get a bit of conscious knowledge about it. When you keep that in mind while coming across the word again in future, it should help you understand the wider context, and if so, your unconscious mind will have an easier time working on it. Note that it's not the conscious knowledge that helped directly - it's that the conscious knowledge allowed you to understand the context, which is something much more easily digested by your mind. It's similar to how you can not understand a joke, think about it consciously for a moment, then understand it and immediately start laughing. The conscious thought brought forth the context, and the context was understood.

If you're interested in learning more, check out Stephen Krashen's seminal text on input hypothesis, and What I've Learned's and Matt vs Japan's videos on the same subject.

  • 1
    Also, I had no idea about your "big green tree" example, very interesting, I was about to ask what's the rule there but I saw the link posted by @chrylis -cautiouslyoptimistic- in a comment to Draconis' answer and I guess that's the rule. I'm ready to bet that at least 90% of high level but not native English speakers have no idea about it and yet to a native speaker's ear it would immediately sound incorrect even though the native him/herself might not even know about the rule. So interesting. Commented Aug 9, 2020 at 8:04
  • 2
    There is a pop science book about exactly this, but I can't for the life of me remember the name. Anyway, the idea is that all languages base on a kind of universal grammar, that's the hard-wired bit. The author analyses many languages, and adds examples of people who didn't learn to speak as children, etc. Gottit! The Language Instinct by Steven Pinker from 2007.
    – RedSonja
    Commented Aug 10, 2020 at 8:29
  • 1
    @SantiBailors Actually the adjective order rule popped up in one of my feeds last fall and I did a survey around my office and my friends to see who knew about it. My non-scientific result is that people learning English as a non-native language in India and Egypt are often taught this explicitly as a rule. And zero native speakers had ever heard of it at any level. I also find this interesting, that such unnecessary rules can form and continue to exist. (Unnecessary in that there is no ambiguity if adjective order is changed.) Commented Aug 10, 2020 at 13:52
  • 3
    @SantiBailors "which I suppose is fully dedicated to servicing the survival and reproduction instincts" what? No. This is nature, things aren't "fully dedicated" to anything. They are piles of stuff that aren't consistently fatal prior to reproduction, not "dedicated". Commented Aug 10, 2020 at 14:36
  • 2
    @user3067860 Thing is it isn't an arbitrary ordering rule; we just order the modifiers so that the properties which are most inseparable from the modified word are closest to it. If I'm writing a fantasy novel where 'great dragons' are a distinct species that comes in many colors, then "green great dragon" will be more natural in that context than "great green dragon", because "great dragon" is its own 'lexical entry'. The list in the Anderson tweet is a) overstated and b) just listing of properties in order of their likelihood of changing. We switch the order for emphasis all the time, too.
    – Tiercelet
    Commented Aug 10, 2020 at 19:13

One aspect is that you're suffering from confirmation bias. Those people define their dialect. The only ones to judge the correctness of their speech are themselves. Everyone is a master in their own game. Even if it's external observer (e.g. you) who are judging their proficiency - what is the benchmark to judge them against? Themselves.

The other aspect is that language creation is an innate skill of humans. It's in our genes to create and use perfect (again, as judged by humans) language rules. More shocking is that children are way better at it. Pigdin (incoherent mix of different languages) evolves into creole (a proper language of its own) only after a new generation is born and raised. When learning their first language, children instinctively create valid rules for it and discard the conflicting parts, effectively creating new language.

The objective of teaching is not to teach children the rules. They will pick them instinctively on their own, and modify to be even better. The objective of teaching is to make sure people from different parts of the country are learning same rules. Teaching basically tries to stop language evolution, to prevent splitting it into mutually unintelligible dialects, which will then become independent languages.

  • 2
    True that those people define their dialect, but it's not that some individuals decide a rule and the others comply with it. My question was about how all the speakers of a certain dialect achieve perfect collective compliance with rules that aren't written anywhere or taught by anyone. The benchmark is how consistently the same rule is complied with across all the different individuals, not how proficient they are in their dialect which as you correctly say I cannot judge. To me the interesting part is the spontaneous and yet collective and perfect compliance with the same unwritten rules. Commented Aug 9, 2020 at 12:47
  • 2
    @SantiBailors I guess we still can't look at the issue from the same point of view. My point is that exactly it's not something that some decide and others comply, but it happens spontaneously as a compromise worked out by people using it. It happens on the internet all the time, when one person says "k" instead of "ok" and we all think "I want to be as cool as this guy" and we all start saying "k" without any agreement or rule-setting body. We just imitate things we consider good, and new standard is worked out this way, without us even realizing it's a new language rule we've just created.
    – Agent_L
    Commented Aug 9, 2020 at 16:01

It comes down to the difference between natural acquisition and book-learning. Everybody learns their native language perfectly, just by constant exposure and use. Nobody ever teaches you the rules of your native language, you induce the grammar from observation of language use. (Well, sometimes people try to teach other how to talk their language, but it is completely ineffective). When the standard language isn't actually spoken in a particular area, and you have to learn it at school, then it's basically a foreign language. Some people learn foreign languages in school well, others do not. Illiterate people tend to not go to school (that's where you learn to write), so pretty much by definition, illiterate people are those who don't learn to read and write the standard language.

  • 1
    Great point about the foreign language. It seems quite obvious now. Basically if I learn a language as MY native language I spontaneously comply with the rules. Commented Aug 8, 2020 at 19:16
  • 3
    You do, @SantiBailors, But unfortunately there is a long tradition of people in power deciding that the way they speak is "correct" and any other way of speaking is "incorrect". So for centuries people (especially children, but not only children) have been subject to the systematic abuse of being shamed or punished for speaking their own language perfectly.
    – Colin Fine
    Commented Aug 8, 2020 at 21:03
  • 2
    @ColinFine You make it sound like that's an arbitrary decision instead of having some practical reason for wanting a standardized language. I'm not trying to excuse repression or discrimination of minority languages, but it actually is in the interest of a ruler that all those they rule can communicate effectively with each other, and it's arguably in the best interest of those they rule as well. Commented Aug 9, 2020 at 16:34
  • 2
    @AustinHemmelgarn, the practice of actively suppressing a person's native language to encourage monolingualism in the standard dialect is what is objectionable. Nobody objects to teaching of a standard language to "empower" citizens.
    – user6726
    Commented Aug 9, 2020 at 16:51
  • @AustinHemmelgarn I totally agree with your point. In my interpretation, Colin Fine's criticism was right too because it sounded aimed not against having a unified national language, which to me is a great thing, but rather against decisions to forcefully change the rules of a language from above despite the way it has been commonly spoken up until then. It reminded me of Mussolini's attempts at some such forced changes, which of course failed. I think France too sometimes tries to forcefully direct the language, even with laws IIRC. But promoting a homogeneous national language is good. Commented Aug 9, 2020 at 16:53

As the famous saying goes, "A language is a dialect with an army and navy" and whether or not something is a language or dialect depends on historical circumstance e.g. Galician is considered a Spanish dialect, Catalan isn't, but doesn't have it's own army and navy(yet) and Portugese has it's own army and navy(and air force). From what I've heard about mutual intelligibility from Italians(a guy from Milan, to be precise), he was pretty lost when it came to understanding colloquially spoken Sardinian in some back-woods village.

If you think of the situation as a first-language Portuguese speaker in some alternate history where it doesn't have it's own army and navy, sounding weird when speaking (Castilian) Spanish, then the situation won't seem strange at all.

  • 1
    Nitpick: these days Sardinian is considered a separate language, not a dialect of Italian. But the same holds for many other Italian dialects, though. Commented Aug 9, 2020 at 8:31
  • Thanks, it's actually not so clear to me what properties must be present in a dialect for it to be considered as a dialect. The description of the "dialects" tag here requires mutual intelligibility between it and other dialects of the same language and this would automatically make "non-dialects" all the "dialects" of Italy I am aware of (e.g. Sardinian, Neapolitan, Venetian and hundreds more), because no mutual intelligibility is required or expected and usually it's not present. I will have to read on this subject, and @FedericoPoloni's link in the comment above seems a good place to start. Commented Aug 9, 2020 at 8:38
  • @SantiBailors I don't think that is a clear, well-defined topic at all. (Just for fun, though, I'll note that in the past a Sardinian army and navy have existed, so this fits Eugene's criterion) Commented Aug 9, 2020 at 8:44

Already many very good answers, so just one additional aspect.

Have you ever been asked by a foreigner who was learning your native language about some grammar rules? I bet you won't be able to answer (unless you have been specifically educated in teaching the grammars).

It happened to me, a native German speaker, and the question was: "Is the third person plural of a German verb always identical to its infinitive?" He had found that rule in a textbook and wanted confirmation from me. The best I could do was imagine some example sentences and then compare the verb forms, but of course that didn't help him in his rule-based language-learning model.

As a native speaker, you don't think in categories of rules when speaking, you just know what words and forms to use. Learning languages by rules is un-natural.

  • I totally agree, and I find fascinating how native speakers all spontaneously comply perfectly with the same rule regardless of whether they have ever read them or had them taught or even know they exist at all. Commented Aug 10, 2020 at 16:16
  • No (sein). Presumably that same textbook had a table of irregular verbs :-) Commented Aug 12, 2020 at 0:34

Previous answers and comments have clarified the issues around your questions, but I offer Standard Chinese (putonghua) as an interesting example of the issues. It is based on - but not the same as - North Chinese dialects, and its rules were codified by the Nationalist government in the 1930s and the Communist government in the 1950s. Thereafter it became the official language and was taught in all primary and secondary schools throughout China. It is now common for Chinese citizens to use dialect for their in-group and putonghua for communication with government officials, businesses, and strangers.

The result is that, while 100% of "native Chinese speakers" are perfectly competent in their native dialect, very few of these dialects have written forms, so literacy is defined in terms of competence in putonghua. Perhaps as much as 20% of the population writes putonghua poorly or not at all, mostly the elderly, ethnic minorities, and people in remote rural areas. (Official stats claim that literacy is almost universal.)

My point is that linguistic competence is a gradient with several aspects, rather than a black/white distinction. The lines between dialect and standard language are not only blurred, they also change over time as current usage changes. A clear majority of "native speakers" are "perfectly fluent" in putonghua because they studied it from the age of 5 on, but they still retain distinctive accents and they often import dialect elements into their use of putonghua.


Minimising mental effort

One of the reason a language develops precise rules in the first place is so that you don't have to think about how to say something, and a listener doesn't have to think hard to understand it. The simplest example is resolving ambiguity, e.g. using word order (or inflections) to distinguish "Paul followed Ann" from "Ann followed Paul." Poetry often uses unusual ways of saying things which force you to think about it (which is probably the point.)

Our brains are pattern-matching machines and if an idea can be expressed by plugging a few things into a standard "template", and then understood by the listener by pattern-matching against common templates and stock phrases, it takes less energy and time to choose words and to understand them.

When we say things that don't quite follow the "rules", we sense that the listener is a little bit uncomfortable with this, which helps us adapt to the rules. This is much easier for children.

Of course, it is all about incentive. Your brain cares most about rapid, fluid communication with people that you meet every day, and that is ultimately the answer to the question.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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