On the English Language Learners SE meta site, I'm involved in a conversation that I find somewhat bizarre. In the order in which they appear, here are the assertions (paraphrased with supporting quotes) made by the user with whom I started the conversation:

  1. The distinction between hyphens and dashes in English is not part of English or any other language: "...dash style is [not] part of language use, whether English or any other. Of meeting some formal style prescriptions, okay, but not language qua language."
  2. Capitalization is also not part of a language: "...they're not part of language, they're just stylistic, like capitalisation.... 18th-c. writers in English capitalised nouns. The style changed...but the language was unchanged."
  3. The distinction between hyphens and dashes doesn't affect understanding: " Because, like capitalisation, they don't affect understanding. They might affect us on a social level... but the reality is that our ability to communicate is unaffected.... which dash we use, or whether we use a dash at all, doesn't alter the meaning of what we're expressing..."
  4. The written word is not part of a language, but merely a "representation" of spoken ("real") language: "[The written word is] a representation of the language, but it's not the language. It's comparable to a map. Maps aren't the territory, they just represent it."
  5. It's impossible to know a written language but not know the spoken language: "Plenty people are illiterate in a language they know well, but you'll never find the opposite situation."
  6. Not only is written language not a part of the language as a whole, it's also not a separate language, so it's apparently not "language" at all: "...while you can call a written representation a separate language, you'll get little support from linguists (or polyglots)."
  7. Some punctuation marks "have a role" (?), but not hyphens and dashes, which cannot possibly affect understanding: "Commas, periods, and semicolons have a role because their use can actually change interpretation. But an en dash vs a hyphen? No earthly way."
  8. Although written English (per the above) is not a separate language, it certainly seems like one: "I sympathise with your feeling that written English is a different language. Many non-native speakers suspect it comes from a different planet. English spelling conventions are as bad as Gaidhlig's."
  9. Even if written language is considered a "dialect" of the English language, the hyphen/dash distinction still wouldn't count as part of the "language", because it's "essentially impossible" (?) to distinguish between them when writing by hand: "For one thing, it's essentially impossible to get the 'correct' length when writing by hand. Yet if text-mode English is to be called a separate dialect, then handwriting must be an example of it."

My impulse is to disagree with every single one of these (except the point that written English has strange spelling conventions), but I have not formally studied linguistics. So while I can provide counter-examples of some of the claims that are easily analyzable (e.g. I've provided examples in the thread of statements that could cause confusion if dashes are replaced with hyphens), I don't really have the background to adequately address the more abstract or definitional claims with regard to how "real linguists" would view them. However, the user who made these claims also lacks the appropriate linguistics training to support them. (We've both admitted this ignorance in the thread.)

So, can anyone here support or discredit these ideas, either by appealing to more rigorous and well-established definitions of the terminology involved, or by presenting some other evidence of how linguists approach these issues?

  • 2
    I think there are 2-3 separable questions here. E.g. "Are hyphen and dash semantically different?", "Can a language be written?", and the title question about English, as stated. The list item I find most surprising is #5. I must be misunderstanding it, because every instance of studying a language [especially a dead one] from documents alone would seem to be a counterexample.
    – WAF
    Commented Jan 11, 2017 at 19:20
  • @WAF The user is not unaware of dead languages, and even brings up some examples I hadn't heard of. Please feel free to look through the transcript and let me know if you think I've misunderstood them. There's even a counterexample of another user who joins the thread and describes learning written English without learning spoken English, but this appears not to influence the other user's opinion as far as I can tell. Commented Jan 11, 2017 at 20:18
  • 3
    anybody who thinks punctuation marks are part of any language will have to explain how illiterates manage to speak a language.
    – mobileink
    Commented Jan 11, 2017 at 21:02
  • 1
    @mobileink Actually, your comment sheds some light for me on what might be a fundamental misunderstanding. By "part of some language," I never meant to imply that it was an indispensable part of the language in the sense that the language couldn't exist without it, or that people who don't learn written language don't "know" their native language. But I would also consider obscure vocabulary to nevertheless be "part of" their language, despite the fact that very few speakers need to know those words. Commented Jan 11, 2017 at 22:07
  • 1
    (Note that in the original context, "part of the English language" simply meant "relevant to the study of English by English language learners as facilitated by ELL.SE.") I think that written language is fundamentally linguistic in nature. My understanding of the other user's position in the linked thread is that they do not agree with this statement, but perhaps I am misunderstanding them? Commented Jan 11, 2017 at 22:08

3 Answers 3

  1. The use of hyphens and dashes could be considered part of a language's orthography, literally its "correct writing". Orthography encompasses everything from the alphabet to spelling to capitalization to punctuation, and is certainly a part of language usage. But whether it's an intrinsic part of the language itself is up for debate. Some languages have multiple orthographies, and spoken language can be transcribed in a phonetic notation without any reference to the orthography, but you might also plausibly be reading this English text with no knowledge of its spoken pronunciation.

One idea about this is that there are multiple "levels" of language: there's the ideal Platonic representation of English in your mind (langue), and then at a different level it can be represented as written text or audible speech or Shavian transcription or... (parole). This is a very clean and elegant analysis, but there's also evidence that spoken and written language are different levels: infants learn to speak (or sign) significantly before they learn to write, for one. So written and spoken language are not entirely equivalent.

  1. Capitalization is also a part of orthography. German orthography for instance specifies that nouns must be capitalized, and also that the phoneme /x/ is written «ch», and that «ß» is a part of the alphabet, and so on.

  2. The distinction between different types of dashes can sometimes be important, but people who are not typesetters usually don't care.

  3. As a counterexample, written language can certainly affect spoken language, through phenomena like hypercorrection (when something changes because people think it's wrong, even if it isn't). The «th» in "clothes", the «l» in "falcon", the «h» in "forehead", the «k» in "ski", and many others used to be silent in English. But due to the spelling, people now usually pronounce them. Using their metaphor, you wouldn't expect a misleading map to actually cause changes in the territory, with lakes appearing suddenly due to a blue spot on the map.

  4. I can read Ancient Greek, but my pronunciation is atrocious and I doubt I could find a toilet in ancient Athens. To a lesser extent, I can read the difference between aspirated and unaspirated consonants in Hindi far more easily than I can hear it.

  5. Usually written English isn't considered a separate language from spoken English, for instance. Whether or not the orthography is truly part of English can, again, be debated, but even if it's separate, it's not a language in and of itself.

  6. Most people wouldn't notice the difference, especially in the age of ASCII text.

  7. English orthography is indeed an enormous mess. Usually there are historical reasons for the weirdnesses, such as the Great Vowel Shift, but this is little help to a child trying to memorize irregular spellings.

  8. Handwritten text tends to be a bit sloppier than typewritten text in certain ways. The precise width of different dashes is one of them. I very much doubt they could be distinguished in my handwriting, or that of most people I know.


Being a linguist, I agree with all of them except for #5 (which is untrue). Written language is a late-developing technological development -- through most of its evolution, language has been spoken only (or sung).

However, I'll suggest to you that it doesn't matter, from a scientific perspective. Scientific theories are generally taken to define their own subject matter. If quantum mechanics is found to provide an explanation of the origin of galaxies, no one will object that galaxies are the proper domain of astronomy and not of physics. The explanation comes first.

So, if it should happen that the next big linguistic theory gives a central place to punctuation, linguists will have to change their minds and concede to writing a less peripheral place in the analysis of human language.

In the meantime, if you look at the theories of human language currently under discussion, you'll see that there is no appeal to writing at all. So far, there is no support for writing playing any essential role in human language.

  • I think agree with your statement that it doesn't matter scientifically--in general, definitional arguments don't really have scientific meaning. In context, the original fundamental question (which isn't quite appropriate for Linguistics.SE) was whether or not discussion of hyphens and dashes belonged at all on ELL.SE, with the user's claim being that they don't belong because they're "not language." So, to me, the corresponding "fact-based" question is, "do English language learners generally benefit from learning to distinguish hyphens from dashes?" Commented Jan 11, 2017 at 20:22
  • 1
    Either way, however, I'm not sure how the fact that written languages developed much later than spoken languages bears on any form of the question. Just because written language came after doesn't mean it's not part of the same category, in the sense that it shares the same relevant characteristics (used for communication between humans, syntax-based, etc...). Commented Jan 11, 2017 at 20:26
  • I'm also not sure why "written language counts as language" would imply "theories of human language... appeal to writing." My opinion as stated in the thread and implied in my question is that written language "counts" as language, not that it is an inextricable/inalienable/indispensable part of all languages (which would be an absurd claim). Commented Jan 11, 2017 at 20:26
  • @KyleStrand: In context, the original fundamental question ... was whether or not discussion of hyphens and dashes belonged at all on ELL.SE, with the user's claim being that they don't belong because they're "not language." I think you must have misread me, Kyle. If I thought that they don't belong in ELL, why did I suggest the addition of a "dashes" tag to complement the "hyphen" tag? Language use affects one's place on the social-status continuum, too, and knowing how to adhere to convention, whether in dash use or fork use, is important to the extent that social status is important.
    – MMacD
    Commented Jan 13, 2017 at 13:39

As you're already aware, in one sense, this could be said to be just as trivial as any other terminology issue. Some people use one definition, other people use another, and it doensn’t really matter as long as you know what definition each person uses. I think it's silly to argue that English language learners cannot benefit from questions and answers about orthography. The way linguists define language has to do with certain ideas about "linguistics" as a scientific discipline, and what its proper object of study should be; I don't think it's very relevant to people who aren't trying to study language scientifically.

I think to linguists, the most important characteristic of "language" is its universality: in every known human culture, language exists and is universal among healthy adults. All healthy children successfully acquire language somehow.

Speech is related to it: everyone agrees that speech is somehow related to “language”. Writing is obviously related to speech, and since I think “related to” is transitive, I think we can actually say for sure that writing is related to language.

But as you’ve noted, some people don’t like to say that writing “is” language. Why might this be? Well, for one thing, writing is very obviously not universal. Also, as far as I know no child acquires writing before acquiring language in some other modality. Obviously adults who already have acquired language can learn how to read or write some other, non-native tongue without learning how it sounds (although my impression is that empirical inquiry shows this to be a sub-par strategy for gaining competence), but that's irrelevant. Now, this doesn’t actually prove that writing can’t be part of language. But if language is universal and writing is not, it follows that the existence of language does not depend on the existence of writing, or come from it at the level of the individual. Pretty much all evidence also points to language being older than writing historically.

Another reason for stressing the unimportance of writing to language might be to counter a misconception that I think is thought to be common among laypeople: the idea that speech is a debased or degraded derivative of writing, or that writing is closer to language than speech is. (If people do tend to have this idea, it might be because it’s common in literate societies for individuals to encounter words in their written form before ever hearing them spoken.) And one easy way of stressing that “language does not come from writing” is to say “writing isn’t language”.

Anyway, so I said that it’s important that language is universal, but what is it? As far as I know, we don’t really know for sure (in an interesting way, beyond the obvious). We know bits and pieces, but not the whole thing. That's why there's such a diversity of theories. We can get data about language from both speech and writing. Some theoretical reasons for preferring speech over writing as a source of data about language is that speech is much more commonly used, on both a personal and a societal level, and it’s the first and most robust thing children acquire (and language acquisition is one of the important fields of study). Some theoretical reasons for preferring writing over speech as a data source is that it’s easier to collect written data than it is to collect spoken data, and writing generally has more examples of complicated structures than speech (some kinds of complicated structures basically never occur in spontaneous speech; although then again, there are also different viewpoints about the value of linguistic data from spontaneous sources vs. linguistic data that is consciously created).

The study of punctuation &c is hardly the only thing that can be distinguished in this manner from the study of “real language”. There are also facts about spoken language that are likely to seem (theoretically) uninteresting to many linguists (= language scientists); which specific facts these are probably differ based on the linguist, but for example, there’s a remark Noam Chomsky made comparing sociolinguistics, which mainly deals with observations about variation in speech, to butterfly-collecting; his point as I understand it is that natural science involves the description of important theoretical principles rather than the cataloguing of facts about the natural word (which Chomsky calls “natural history”).

After looking through the linked comments a bit, a few more observations that I think might help.

You say

languages are nothing but social convention--every aspect of them

I think the idea that language can be meaningfully explained as just "social convention" is somethign that linguists (at least, those of certain schools) would tend to disagree with. Linguists like Chomsky are interested in the "language faculty" or "language module", the thing that is behind the human capacity for language use; what exactly this is is still not pinned down but it is generally believed to have a lot to do with biology, neurology etc. Obviously it's related to society as well but that doesn't mean it can be entirely reduced to social elements.

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.