I am not entirely sure if this is the appropriate site and whether these are the appropriate tags for this topic, so if that is not the case, please feel free to let me know.

I never thought twice about how reading and writing works until recently and wondered whether how I think it works is actually correct. The way one usually learns reading requires that one previously learns the spoken form of a language. That way one has already attached sounds to a meaning, for example the sound /kæt/ (of cat) to the concept/to an image of a cat. Now, strictly speaking, words are only strings of symbols. That is, the word "cat" is only a string of the symbols "c" "a" and "t". Each of these "atomic symbols" are called letters and are associated with a certain sound. By combining the sounds of the letters that make up the word, we end up with a sound. If we are familiar with the semantics of that sound, that is, with the meaning of the word, we associate a meaning with that word. So when reading the string "cat", we associate with it the sound /kæt/ and with that sound the meaning of a cat. This is basically how I have always read - I translated the symbols/strings to sounds and those sounds have a meaning. That way I know what these symbols try to tell me. That is how I think reading usually works.

Writing is sort of the inverse process. We have some words/sentences that we want to express. These are usually in the form of sounds. To stick to the cat example, without spoken language I would need to point to a cat and make the sound /kæt/ and hope that the person I am trying to communicate to understands that I want to attach the sound /kæt/ to the concept of a cat. Now, when one has learnt the spoken language this is done already. So when one wants to communicate some phenomenon, one expresses that phenomenon into sounds and these sounds into symbols/letters. These are exactly those letters that when read produce the sounds of the thing I wanted to communicate. In the cat example, I know that the string "cat" is pronounced /kæt/, so when I want to write this down, I need to go exactly through this process.

Now what I find fascinating is that all of this happens unconsciously now that we have learnt this and that probably everyone takes this for granted. The processes I describe above are not trivial in my opinion, which makes this pretty impressive, despite being used every day.

Having thought about this and having a mathematical background, I am particularly interested how true statements and logical consequences are communicated. Consider a simple statement like "Today I woke up at 7 am". I already used the word statement, however at first this is only a string of symbols. Trying to read these symbols I get a sound which associates the meaning that I woke up at 7 am today. I can now ask whether these symbols give a true statement, but what does it mean for these symbols to be a true statement? I guess it means that the spoken statement associated with the string of symbols is true, and this in turn means that the meaning of the spoken statement is true. So when communicating facts in written form, we have a phenomenon, sounds that describe that phenomenon, in this case the sound associated with the sentence "today I woke up at 7 a.m.", and a way to turn these sounds into symbols/written form. By the above, the written symbols are true if and only if the phenomenon it is associated with through the sounds is true. This means that when judging whether a written statement is true, we translate it into sounds, check if what the sounds represent is true and then we know whether it is indeed true or not. When trying to deduce things logically from a written statement such as "today I woke up at 7 a.m." one checks which semantic implications these symbols have. This is done by again, translating these symbols into sounds, then these have a meaning and this meaning may imply other meanings, for example the meaning of "I had enough time to shower at 7:05 a.m.". In summary, one would say that the written symbols "Today I woke up at 7:05 a.m., so I had enough time to shower at 7:05 a.m." are a true statement, because the meaning associated with the sounds of the symbols is true. In particular, someone reading this statement other than me could check the truth by translating the symbols into meaning and checking whether the meaning is true.

So in summary, how deduction with written premises works is, one translates these premises to sounds and these to meaning, which may imply other things to be true as well. These implications can be turned into sounds and then into written form again.

I would like to know whether what I came up with regarding this topic is somewhat correct or completely wrong and gain more insights into all of this, which is why I would appreciate any links to further material. Since I am not a linguist at all and previously never really thought about any of this consciously I am sorry if I am imprecise in some sections. Again, feel free to ask for clarification if anything is unclear. Thanks in advance!

  • 3
    You're right to question the common assumption that written language is primary: it is not, and is a more-or-less effective way of recording and transcribing spoken language. But just because it is historically contingent, that does not necessarily imply that written language must remain that way. It is certainly true that nearly everybody learns to speak or sign their native language before they learn reading and writing. But many who have studied classical languages such as Latin were barely able to speak a sentence of the language, and did indeed evaluate the truth of written sentences.
    – Colin Fine
    Jan 7, 2023 at 20:28
  • @ColinFine I wasnt aware that it is a common assumption that written language is primary. Why should that be? I find that quite unintuitive. Regarding your latin comment: While I agree, they learnt another language beforehand and can translate the written statements into written statements of their mother tongue, and those again into sounds interpret these. So that doesn't really contradict what I thought, does it?
    – user3118
    Jan 7, 2023 at 21:13
  • 1
    I don't think that people who have any real understanding of language think that, but many untutored people (act as if they) think that. For example, when they want to ask something about sounds, they will talk about "letters"; or they will have a question about punctuation, and think it is about grammar.
    – Colin Fine
    Jan 7, 2023 at 21:18
  • As for Latin: yes, it's possible that people do as you've said. Have you any evidence that they do? Many people who have learnt a foreign language have reported the experience of, at some point, no longer needing to translate into their own language to understand. Why should that not be equally true of people who have learnt a language almost entirely through the written form?
    – Colin Fine
    Jan 7, 2023 at 21:19
  • @ColinFine Thanks for taking the time to give me some insights by the way. I dont really know people that are interested enough in this so that I could exchange about that on. As for latin: sure, I think youre right about that as I would say that I also experienced that. Of course, I cant prove it, since it might be happening unconsciously that I translate it in the way I suggested. However, a translation process still remains. As far as I know, it is even possible to not read with a voice in your head, so that way you also skip a step.
    – user3118
    Jan 7, 2023 at 21:33

1 Answer 1


Many people can read and write a language, having no experience whatsoever with the spoken form of the language, or any language. Sound is not essential to language acquisition, all that is required is some means of symbolic representation (Braille, gestures). I am not sure if any person with normal hearing has ever acquired knowledge of written languages like Chinese or Egyptian without any notions of pronunciation whatsoever, but very many people gain reading knowledge without much grasp of spoken language.

Translation to sounds is completely unnecessary, what is necessary in understanding the "meaning" of a sentence is the ability to grasp specific concepts expressed in the linguistic signal. A sequence like "Those were snakes" embodies a bunch of concepts, like "snake", plurality, deixis and time-reference. In spoken language this is conveyed via particular sequences of sound, in alphabetic written languages by sequences of graphic symbols which happen to have a (more or less) computable relation to sounds but can just be treated as arbitrary drawings (if you don't know the pronunciation of the letters), and can also be conveyed by holistic drawing of some kind – hieroglyphs / logographs.

I personally think it is a mistake to talk about meaning in terms of truth, because only propositions are true or false, and not all meaningful language is propositions. All meaningful constructions have some kind of symbolic representation, perhaps visual, auditory or tactile (our anatomy is not refined enough to have an effective olfactory communication system). Checking truth values presupposes a shared conceptual systems between parties, not a shared system of external symbolic representations.

  • 1
    I find your first sentence to be a bit much and only applies to scholars and dead and/or obscure languages.
    – Lambie
    Jan 8, 2023 at 1:50
  • 4
    I am surprised that you are unaware of the existence of deaf people.
    – user6726
    Jan 8, 2023 at 1:52
  • I think very few people (proportionally) can read and write a language with no experience whatsoever with a spoken form of the language. Very, very few people are born that profoundly deaf, roughly 0.05%. (However, this does not affect the thrust of what you say, of course.) Jan 8, 2023 at 7:12
  • @user6726 I agree with you. I fear that when writing my post, I was too focused on how one usually learns a primary language, although this does not represent my entire view of the topic. Technically, it should be possible to establish a primary language purely with written language and to communicate its meaning by gestures. This may lead to a lot of difficulties, but I would think its technically possible. You also bring up an interesting point I thought of, but had no answers to, namely how deaf people learn languages and whether that is based on something. I should research that a more.
    – user3118
    Jan 8, 2023 at 11:14
  • 1
    @ColinFine Yes, you said it much better than I did.
    – Lambie
    Jan 10, 2023 at 15:25

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.