I have noticed that on EL&U some people seem to take it as axiomatic that only spoken language is real language. This does not seem axiomatic or otherwise obvious to me. Is it just a currently fashionable doctrine in linguistics, or is there an adequate argument in support of it, or both?

Some examples are here

Colloquial English means English as it's spoken, which is the real language. What's sometimes called "formal English" is simply stuffy, an attempt to talk like our betters.

—and here

I'm sure it's already far more common in real (spoken) language.

  • Please provide examples. I know where you're coming from, but folks around here ‘wanna’ see hard evidence.
    – Mari-Lou A
    Commented Jan 5, 2016 at 20:05
  • Thanks, @Mari-LouA—I have now done so. Commented Jan 5, 2016 at 20:30
  • 15
    It is a common linguistic view. The main basis for it is that spoken language comes before written language, both in the life of an individual and in overall historical development, and spoken language is universal among healthy human beings living in human societies while written language is only used by some members of some societies.
    – herisson
    Commented Jan 5, 2016 at 21:07
  • 2
    I agree, @sumelic, that the story of Tarzan's acquisition of English via books first is implausible. (I think Derrida may have claimed that written language was prior to spoken, but I do not pretend to understand what he might have meant by that.) But surely it is the principle that the earlier is not therefore more genuine that allows us to call the etymological fallacy a fallacy. As for universality among humans, all healthy humans possess hemoglobin while but few possess platinum, yet platinum is not on that account thought to be less real. Commented Jan 5, 2016 at 21:24
  • 6
    If you could add up all the words spoken and heard by all the Anglophones on the planet in a year, I'm sure they would be orders of magnitude greater than all the words written and read in that same period. That's over and above the fact that until a century or two ago, hardly anyone could read or write anyway (but all native speakers have always been able to use their own language perfectly well). @sumelic's point is well-made, and I would add that linguists are well on their way to being considered scientists - which grammarians (focused on written forms) never will be. Commented Jan 5, 2016 at 22:31
  • 2
    Physics is universally held to be a science, and literary criticism to be nothing of the sort; but Wayne Booth justly pointed out that we can nonetheless be more confident that the opening sentence of Pride and Prejudice is ironic than that the laws of physics as currently understood will still be standing as such next year. In other words, being an object of a kind of study or understanding that is regarded as science does not ipso facto guarantee superior ontological status or even certainty. Commented Jan 5, 2016 at 23:44
  • 4
    On this topic, I highly recommend the introduction to Geoffrey Nunberg's The Linguistics of Punctuation (1990), starting on page one. It's titled In Search of the Written Language.
    – user28567
    Commented Jan 6, 2016 at 0:38
  • @BrianDonovan I think there is some truth in the pride and prejudice observation, but I also think it stands on shaky ground. Will people still think, in 400 years' time, that the opening line is ironic? What about Shakespeare, Homer, to give counter examples? And yes, I agree that science has done a great PR job in making itself seem the arbiter of truth, and everything else as subjective, and therefore "false." Commented Jan 6, 2016 at 2:01
  • @BrianDonovan may I ask what prompts you to ask your question? Does it seem to you that answers are given or accepted based on spoken superiority and that this needs critical examination? (Personally, I am not convinced that spoken language is the true form of the language.) Commented Jan 6, 2016 at 2:04
  • 1
    @michael_timofeev I don't know where you've been getting your science "PR", but science makes no pretense of being the "arbiter of truth." Science is the arbiter of validity among various models built according to its rules. Scientists agree that all currently accepted models are only provisional, contingent on the next discovery.
    – deadrat
    Commented Jan 6, 2016 at 2:36
  • 2
    This debate should perhaps be migrated to Linguistics. Commented Jan 6, 2016 at 2:58
  • 1
    @michael_timofeev I suspect that the word doesn't mean what you think it does, especially considering that you can't spell the man's name. In any case, there's no need to appeal to mathematical logic to find statements that science can't prove. Science doesn't prove things at all. That's for mathematicians, bakers, brewers, and minters. Science validates models, none of which deal with ethics, esthetics, or the teleological.
    – deadrat
    Commented Jan 6, 2016 at 4:05
  • 1
    See the same question on linguistics.SE for some more perspectives.
    – Mitch
    Commented Jan 6, 2016 at 22:56
  • 2
    "Only written language matters" and "only spoken language rocks" are two opposing, and largely nonsensical, views.
    – Ricky
    Commented Jan 7, 2016 at 1:22
  • 1
    One clue should probably be that changes in the spoken language being reflected in writing is much more common than the other way around. Can you think of some grammatical form which was created by being written and then became widely accepted in natural speech?
    – Casey
    Commented Jan 12, 2016 at 21:28

6 Answers 6


Even supposing that the putative primacy (it comes first) and commonality or frequency (it comes more often) of spoken language can be established, those traits don't confer 'ontological superiority', that is, they don't make spoken language any more or less 'real' than recorded language.

The view that primacy, commonality and frequency somehow make spoken language more 'real' should be more accurately expressed as the "currently fashionable doctrine in linguistics" that those characteristics make spoken language a fitter, more viable subject for contemporary linguistic studies than recorded language. That expression of the case, then, as opposed to an ascription of greater or lesser 'reality' for spoken versus recorded language, recommends a type of "ontological superiority" that doesn't set up and burn a straw man in one smooth flick of a Zippo.

An ontological feature of humanity is the use of language. ... Different from the transient nature of spoken language, the recordable character of the text significantly developed the brain, social life, and intellectual skills of humans.

("Superiority of Hangul" at The Duksung Vista, 20101004.)

Although the propositions in that excerpt are deployed in the service of what must be considered a parochial argument (the superiority of Hangul), in isolation they are evidently true, and speak to a distinctly different view of ontological superiority than the aforementioned "currently fashionable doctrine in linguistics". In this view, the 'reality' conferred by comparatively greater and more recent contributions to human development assumes an equally greater and more recent ontological significance.

One such human development is the study of linguistics, which itself is only possible by means of recorded language. To neglect this, that recorded language is the medium and sine qua non of linguistic study, in favor of a blinkered view that posits spoken language as the only or the more important subject of linguistic studies, is to neglect the essential nature, as well as the necessary philosophical underpinnings, of linguistics as a scientific endeavor.

Thus, I contend, the "currently fashionable" linguistic doctrine that argues the primacy, frequency and commonality of spoken language make spoken language a fitter more viable subject of linguistic study must yield to the primacy, frequency and commonality of the language of linguistics itself, that is, recorded language.

  • 3
    Yes, but recorded language isn't equivalent to written language. These days audio and video recordings are commonly used for linguistic analysis. :) Commented Jan 6, 2016 at 2:58
  • @curiousd - I think what you say is closer to the point. It's not that one is "superior" to the other, it's apples-and-oranges. If we are discussing spoken language down at the pub, we need to be careful about using ngrams to support our stance, just like we shouldn't we use random internet message boards to help us choose language for a technical report. I think the OP is slightly misinterpreting the cited comments by saying they imply one language is "ontologically superior" to the other. It's different (and the differences are noteworthy) but not necessarily "superior" in either direction.
    – J.R.
    Commented Jan 7, 2016 at 9:56
  • 3
    But by this logic, you can't study chemistry without written language either, so should chemists stop studying chemicals and reactions? No. Science without written language isn't impossible, just very inefficient. So it doesn't follow at all that written language should be the object of the science of language. And there a plenty of ways in which written language is the only source of data we have: historical linguistics in one such area. So it's not that linguists don't ever study written language.
    – Alan Munn
    Commented Jan 7, 2016 at 18:05
  • @AlanMunn I'd add that in many places or in different time periods one can observe a much bigger gap between what is written and what is spoken, sometimes to the point of the written language being entirely unintelligible, if read aloud, to a fluent speaker of the spoken language. I think that parts of the world where the written language is Chinese or Arabic are the obvious modern-day examples.
    – Casey
    Commented Jan 12, 2016 at 21:32

It was a common view in the 1800's (especially continentally) that the written word was ... better than spoken. But that is an inchoate comparison. And the common view now is that spoken is 'better'. To answer your question we would need to specify what the criteria are. 'Ontologically superior' is marginally more meaningful than 'better'.

So let's try to specify the decision criteria.

Obviously if the question is about pronunciation, then spoken wins over written (especially in English). So we can easily dismiss that question

What leads to clearer communication? I think it is obvious that writing has less room for misunderstanding, at least in the transfer medium (one can equally confuse people with the choice of words an argument in either).

What is more lasting? Again, it is obvious that the technology of writing is much more permanent than speaking.

I could continue to list all the ways where writing is better than language. But I know what your question really intends. Why do people on ELU say things like spoken language is real but written language is not?

On ELU we are asking and answering questions about communication in English. We are using writing to communicate these ideas, not speaking, but the subject matter is the language itself. And the questions are of the form 'Is it ok to say X?' or 'What is the word for Y?'. So the real question you should be asking is what the criteria for judging such questions are.

What is in writing that is not already in spoken language? The alphabet or character list. Anything else? Oh, orthography. In English that is complex, but the complexity is ... not linguistic. Spelling is a poor attempt at capturing sound, not the other way around.

The OED is a written artifact and it is created to record the meanings and histories or words in English. It's data and support for entries can only be written samples. Writing is a record of the way people speak. Speaking is not a record of the way people write.

It is an understandable confusion, especially for those learning a language. It is much easier as a second language to learn from written accounts, simply because of the convenience of the medium. You don't have to have a native speaker always at hand when you can just look up in a dictionary or search online.

Also, though it is unphilosophical to refer to reality, I feel it necessary to bring up the matter of invention, arbitrariness and biology.

  • writing is an invention by individuals, a spoken language is a continuous process of a community. Sure, someone can invent a new word that everyone uses, but that would be one out of thousands of existing ones.
  • because writing is an invention, it is arbitrary. Its elements can be changed easily and still be equally functional. Chinese can be communicated in hanzi characters or pinyin romanization. Roman characters transcribe Latin as well as English. Surely there is arbitrariness in spoken language, too, signs and symbols and Saussure and all that. But an alphabet or characters is infinitely simple in comparison to spoken language.
  • Humans have evolved biological processes specifically for processing spoken language, both for listening and speaking, brain areas for processing, and mouth, nose and lung physiology for controlling speech. We didn't evolve fingers for holding pens or for typing.

Also, we've been speaking for tens of thousands of years, but only writing for a few thousand

Writing is a better medium for communication, but speaking appeared first and writing is an attempt at recording speech, not the other way around.

So when someone asks a question about whether it is right to say "I ain't no snitch" vs "I ain't any snitch", the answer is about spoken language and is decided for spoken language, not about written language.

  • Great answer. One question; Can we really say that the written word was an attempt to record speech? This is slightly off topic, but relates to your statement. Commented Jan 6, 2016 at 4:57
  • If we didn't have the written word, today we would be all talking in different dialects, every neighbourhood would have its own unique dialect, we would be divided into tribes, and there would probably be more rivalry and conflict. Lies, murders, theft would probably be on a much larger scale because there would be no written code or laws. Music, theatre, medicine, science and maths would not have survived and continued to have progressed if they had not been recorded. The written word has helped standardized language more than anything else...it is responsible for our civilization. IMO.
    – Mari-Lou A
    Commented Jan 6, 2016 at 5:52
  • @Mari-LouA You make an interesting point about civilization (I don't totally agree; certainly writing makes all the good things easier). I think I made the similar point that writing is a better communication mechanism. But that is not the subject of the OP. The OP wants to know why spoken language has primacy in language questions.
    – Mitch
    Commented Jan 6, 2016 at 14:06
  • @michael_timofeev I suppose that 'writing is an attempt to record speech' is questionable, but the opposite is inarguably false.
    – Mitch
    Commented Jan 6, 2016 at 14:09
  • @Mitch I'm not an early writing expert but from what I know a lot of the earliest "writing" we have is pictures on cave walls and they seem to be mnemonic devices to record or remember events in a tribe's life...I think the idea of writing actually reflecting speech started before cuneiform but got into full swing with the Babylonians and then the Egyptians. Hieroglyphs are a mixture of pictures and phonetic symbols but you can actually read them and it sounds like speech. Commented Jan 6, 2016 at 14:30
  • @Mitch how people made the jump from mnemonic devices on cave walls to writing as a reflection of speech is probably not known with certainty...we might never know. Commented Jan 6, 2016 at 14:32
  • 1
    @michael_timofeev we don't know the history of speech very well because the evidence of past speech is gone. Studying writing to learn things about language is like looking for your keys around the lamppost because that's where the light is.
    – Mitch
    Commented Jan 6, 2016 at 17:01
  • The OP writes: "...some people seem to take it as axiomatic that only spoken language is real language" The OP is asking why do some users seem to belittle the role and the significance written language, why do some users consider written language to be "stuffy" and artificial compared to speech.
    – Mari-Lou A
    Commented Jan 6, 2016 at 21:45
  • @Mari-LouA "The OP is asking why do some users seem to belittle the role and significance of written language" ... in the context of answering questions about language not about everything having to do using language. Interesting linguistic questions are usually muddled by boring questions about spelling that are essentially arbitrary non-linguistic choices. All the interesting part of written language is already handled in the interesting study of spoken language.
    – Mitch
    Commented Jan 6, 2016 at 22:33
  • 1
    @Mari-LouA I continue to repeat that I've made the case that written language is a superior communication method for civilization; that's not the OP's question.
    – Mitch
    Commented Jan 6, 2016 at 22:34
  • From some minimal research, it seems that Derrida, a literary critic and philosopher, not a linguist, and a more contemporary thinker than 19th c, made a case for writing over speech, but it is unclear what his criteria were.
    – Mitch
    Commented Jan 6, 2016 at 22:59
  • I think there are definite advantages to either writing or speech and one can't say one is better...certainly writing helped change the world into a more legalistic place and possibly more fair because of writing and because of writing we can build on the thoughts of previous individuals...on the flip side, the past becomes more fixed in stone...the Troubadors communicated without writing because of their memory, and I think writing has definitely affected our memory abilities. Commented Jan 7, 2016 at 11:24
  • Let me clarify (as OP) that I am not really questioning which is better or more valuable--I should hate to lose either. By ontologically superior I simply mean, as I hope my examples illustrate, more truly real. Commented Jan 8, 2016 at 15:30
  • 1
    I agree speaking does not exactly record writing, but the way words are written can very much affect the way they are pronounced. For instance, some of us take care to distinguish the verbs effect and affect in pronunciation, a few of us actually pronounce "duct tape" with two distinct alveolar stops, and quite a few of us pick up a lot of our new oral vocabulary from writing. So the influence is by no means unidirectional. Commented Jan 8, 2016 at 15:40
  • @Mari-LouA The problem is that your assertion is false. The written word does not prevent dialect drift. Latin was a written language too.
    – Casey
    Commented Jan 12, 2016 at 21:42

When you read, for example right now, you will probably have a kind of feint voice going on in the back of your mind. Perhaps this seems like a dubious concept. It kind of should. There's no actual noise coming off the black squiggly things you're looking at right now. So maybe we need to explore this idea. Let's consider the following:

  • hey, sleigh, day

You may have noticed that those words rhyme. That's extremely weird. What does that mean - they "rhyme". It means that the sounds at the ends of these words are the same. But why do you think about this when you're reading? You're looking at squiggly marks. You aren't hearing sounds. These words seem to not just indicate meaning; they very specifically indicate sound to you (unless, perhaps, if you're deaf). And in a very immediate way. The shapes of writing on its own don't have quite the same echoic effect:

  • enough, bough, through

These words don't have the same effect for you as the ones further above. It may even be that the fact they look similar makes them more difficult to read well.

Because you can speak English, you will also get a sense of rhythm from the words that you're reading - even though this does not stem visually from the squiggle in front of you:

  • clear dream states fly straight green stones
  • Georgie Porgie pudding and pie

The lines above will have a very different rhythmic effect, even though you're reading. And in order to get the rhythm from those lines you need to internally translate them into sound.

Now as it turns out, this does not seem to be an entirely illusory phenomenon. When you read you may physically have increased activity in the auditory cortex - that part of your temporal lobe that processes auditory information. In addition, when silently reading you will be making very small involuntary movements of the muscles associated with speaking (known as subvocalisations).

What does this all show in relation to the current question? Well, it would seem to show that the written word involves the encoding and decoding in some sense of speech, or sound over time. In other words we are translating, to a certain extent, squiggles into the sounds and articulations that we use to communicate with the spoken word.

It may also be pertinent to think at this point about the way that we learn language. All healthy people without some health condition can speak language. The same is not true for written language at all. We learn to speak without any direct tuition, but we need some form of tuition, however minimal, to master the written word. We also need to know some form of language before we are able to start reading. Written language is the encoding of another dynamic language system.

So then, the fact is that is that we psychologically translate written language into spoken language. Although written language has an effect on spoken language, spoken language is more pervasive, more universal. It is in many ways more primary, more biological, more primordial than written language.

Does this mean that spoken language is superior to written language? I don't even know what that means. Writing, of course, has had an enormous impact on human history and culture and our ability to communicate across time and space. Understanding written language is hugely important to understanding human beings and human history, civilisation and culture.

However, I think that what the linguists Brian is citing are angling at is that spoken language is more biologically, cognitively, humanly primordial. It is what we have to psychologically translate writing into to understand it. Spoken language also has a greater impact on diachronic change in written language than the other way round.


You're mis-reading the comparison. It's not "English as it is spoken" versus "English as it is written", it's "English as it is spoken" versus "English as formal rules describe it".

  • Well, yes, but where do the rules come from? Largely from a somewhat idealized version of what a very careful, high-status native speaker would say (if the rules were set once and never changed we'd still be distinguishing between "ye" and "thee" in writing).
    – Casey
    Commented Jan 12, 2016 at 21:35

Agreeing with Mark's answer, but elaborating.

The argument is between being prescriptive or descriptive of the language. The "U" in our name says right up front this stack values the descriptive approach. The only place this really intersects with writing is that written English often has to go through editors, so it tends to have a lot of prescriptive rules applied to it where spoken generally does not.

I'd further argue that trying to be prescriptive with English may have some localized uses, but in general is inherently absurd. The language does not currently have a single prestige dialect to coalesce around. What is considered "Correct English" is very different in England than in the USA (and in Australia, and in South Africa...) Its true that both the New York Times and the London Times enforce style guides on their writers, but they are different style guides.

The best you could do in general is compile a list of what is considered "proper" where, and at some level of detail that just devolves to descriptivism anyway.

  • “The argument is between being prescriptive or descriptive of the language”? Not the way I framed it in the original post! @Mark has opened that door and you have sailed right through it, in a wonderful demonstration of the Maslow’s Hammer effect. Also, by the way, prescriptivists have long been quite at home with the term and concept of “Usage,” as witness the title of H. W. Fowler’s opus. Commented Jan 8, 2016 at 12:51
  • 2
    "Not the way I framed it..." That is indeed my point exactly. You've "reframed" things around writing vs. speaking, when that isn't the root of the issue.
    – T.E.D.
    Commented Jan 8, 2016 at 13:45

Traditionally, speech has been primary and emergent, while written language has been secondary and controlled (by printers like Caxton, dictionary writers like Webster, newspapers and journals such as NewYork or London Times, by universities, such as Chicago and Cambridge with their style manuals, and Oxford and Macquarie with their dictionaries, ...)

Subvocalization back to voice-in-my head spoken language is common but not ubiquitous (in fact speed reading courses teach you not to do it).

Interestingly, social media - kids chatting more by text than voice - is changing this as language change is being influenced by the mores and restrictions of messaging systems. Traditionally, this change would happen in the playground. Similar effects on a smaller scale are happening in the meta language of stack exchange too.

The question of informal vs formal language, and descriptivist vs prescriptivist paradigms, is closely related and reflected in the quotes Brian Donovan chose for the original question. The real point of these quotes is not speech vs text but descriptive vs prescriptive, as others have noted. Some of the early choices, like spelling, were originally designed to distinguish homophones so they weren't homographs, while later language reformers have sought to go back to a more phonetic representation - at the expense of etymology and cognates.

In relation to grammar, the standards have been ill understood and poorly taught. In some cases, Latin grammar rules have been inflicted on us. In some cases, rules of thumb are taught to the poor students (both first and second language learners) that are totally wrong due to lack of insight by the teacher. In other cases it is the student that has misunderstood and overgeneralized or overspecified.

To make this concrete consider some examples - the accusative/dative pronouns and the copula, combined with pragmatic/social/cultural guidelines.

You shouldn't put yourself first ("me and..." or "I and.."). You should use I (nominative) with the copula ("is/am/be"). You should use I (nominative) with conjunction ("and").

These leads to monstrosities like ... on behalf of my wife and I... ... it is I who did that ... (why not "it am I" if it is truly reciprocal)

People have been told not to say "me and my brother", "my wife and me" etc.

The written forms mandate more strongly still the prescriptivist forms that ignorant teachers have been forcing their students to adopt. Upper class speech tends to have incorrectly generalized forms of these, while lower class speech tends to reflect the natural form of the language. Their writing tends to reflect the same differences, but their the application of the rules (correctly or incorrectly, directly or in posteditting) is more likely - so is less likely to reflect their language as opposed to the prescriptivist version taught/learnt at school.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .