Possible situations where this might happen:

  • Early human culture develops in a situation where audio recording and recall is straightforward - perhaps in a place where there is an abundance of tame and very talented parrots.


  • After some catastrophe that kills their parents and destroys most cultural artifacts, a group of children on a distant planet that has become separated from the rest of human civilisation grow up with a smart computer that can speak to them, and record what they say, but has no visual display.

Let's assume that in either of these situations, long-term storage of usefully large amounts of audio is possible, as is searching/indexing the stored audio (and all other things we commonly do with text manipulation, including sending it, versioning it, signing it...)


Someone stumbles upon technology similar to the wax cylinder phonograph before any particular writing system has taken hold. (We can't then assume that there would initially be no issues with search, storage and transport; let's assume that in this case, the technology develops fast enough that these kinks get ironed out before anyone minds too much.)

Anyway, my question:

Would this human culture be likely to develop written language? If so, could we predict any characteristics of that written language?

  • 1
    $\begingroup$ Even in our current world, the long-term storage of useful, large amounts of audio is already possible. What makes you think the written language developed will have any special characteristics, other than being completely undecipherable by us? Take Chinese and English, for example. Both languages are completely different and have different characteristics, but both also derived from a spoken language. $\endgroup$
    – Aify
    Feb 7, 2016 at 8:29
  • 2
    $\begingroup$ @Aify But neither the writing of Chinese nor English initially happened in a world where the long-term storage of useful, large amounts of audio was possible (nor was this possible for most of the time during which those languages have been developing, including their written forms).- so is it it's possible - likely, even - that this fact affected their nature (or maybe even the fact they exist at all in any form like what we see today). $\endgroup$
    – user17965
    Feb 7, 2016 at 8:47
  • 1
    $\begingroup$ No, it's completely unlikely. In order to accomplish the premise you've posed, written communication of some form is necessary. A lot of design work goes into designing a system to store large amounts of audio, as well as the algorithms to search/index said audio, and even that work has to be efficiently communicated between those who work on the system before you can set up your premise. You can't just use parrots either - good luck finding the one tiny section you need in your design when your parrot recites an entire blueprint to you. It's just inefficient. $\endgroup$
    – Aify
    Feb 7, 2016 at 11:13
  • $\begingroup$ @Aify No, it's completely unlikely... How so? surely almost any circumstance can affect the development of a written language? A lot of design work goes into designing a system to store large amounts of audio... - that's been the case in the world we live in now, but have I not dealt with that in my two scenarios (both of which may be unlikely, sure, but that seems to be allowed in worldbuilding!). good luck finding the one tiny section you need in your design... I mentioned in my question that searching would be possible.... $\endgroup$
    – user17965
    Feb 7, 2016 at 12:34
  • $\begingroup$ ...and it's not hard to imagine an audio-based search system - similar things exist now - but at this point it seems that you are mixing in an assumption that the subjects of interest would be the people to design the audio system. This is specifically not the case in either of my two scenarios. $\endgroup$
    – user17965
    Feb 7, 2016 at 12:35

5 Answers 5


I think with that alone, no. Written language would still develop for many reasons.

The first is fidelity. Even if you assure me that your parrots remember perfectly, I will doubt it. Likewise, laying my faith solely in an AI leaves me with little ability to verify records. Also, records naturally include pictures. Those work better with captions, not accompanying audio.

The second issue is portability. Paper moves easier than parrots. Text data is easier to parse than audio, taking practically no processing power. Text is also inherently searchable. Even computers search audio by first converting it to text. This may not be necessarily true with a "super computer", but it is true in reality.

The third issue is human consumption. Generally, people can read faster than they can hear speech for similar levels of comprehension. Even if you practice listening to sped up audio, you will probably top out at about double speed, around 400 words per minute. Meanwhile, it doesn't take much practice with reading to break that.

A fourth issue is the intrinsically different voice and prose that wiring can bring compared to voice. With the most talented writers, this is art. Naturally, this is sometimes done in speech, but they are mimicking writing. No one really talks like that. Beyond this is the value of personal interpretation. Speech makes this harder.

Fifth, there's the value of study. Reading allows you to dart across a page, back and forth, picking up random bits in any order as you interest dictates. This seems impossible to me for speech.

There may be others.

  • 1
    $\begingroup$ On the fidelity issue : Text can be altered as easily as audio, can't it? On portability - I'll edit the question to be clear that we're talking about a world where that is not an issue (along with the aspects I mentioned explicitly, storage and searching). Speed and 'art value' are good points but I'm not sure they would be enough to be a driving force. The "visual random access" point is also relevant - maybe that would be an argument for an ideographic writing developing, rather than (or as well as) phonetic writing that is simply a notation of linear speech? $\endgroup$
    – user17965
    Feb 7, 2016 at 16:42
  • 12
    $\begingroup$ +1 I hate it when I click on a news link, to find that it's a 10 minute video rather than a transcript that I can peruse in 30 seconds and determine whether I want to take the time to read the rest. $\endgroup$
    – Cort Ammon
    Feb 7, 2016 at 16:57
  • 4
    $\begingroup$ Another point is that it's much easier to do semi-random access in written text. Audio is inherently linear, so you have to listen to stretches of it to gain a sense of the content. $\endgroup$
    – jamesqf
    Feb 7, 2016 at 18:17
  • 6
    $\begingroup$ Written documents, even the ealier ones, tend to last longer than parrots so you don't need to keep making copies of your materials. $\endgroup$ Feb 7, 2016 at 18:55
  • 2
    $\begingroup$ 6th, A writer often revises the same sentence over and over again as they develop their thoughts... I do not envy the editor responsible for checking a parrot for grammatical mistakes and unclear references. $\endgroup$
    – Nosajimiki
    Jul 3, 2019 at 20:37

I suggest that there are very strong reasons to develop written language of some form, even with a perfect audio recall system.

  1. Transmission of information. This doesn't just mean computerised transmission, but the wider meaning. With audio, you need to have an environment where you can hear one another. This becomes a problem surprisingly often. Sticking with an an early culture, how do you warn people coming after you about the danger on the trail? If you leave a visual message, you don't have to wait around.

    Getting to a slightly more modern position, how would spies communicate? Even going back to Roman times, secrets have been passed in writing to prevent them being overheard, or to allow the exchange to take place in public. This doesn't work if you don't have some kind of visual representation.

    In an even more modern sense, without writing, how would you warn of speed limits on roads? It's plausible that you could have a speedometer which tells you the speed you're going, but until you've figured out some form of data transmission method, your only hope for speed limit signs is very loud shouting!

  2. Natural environment. When you think about it, humans developed in a audio-biased environment. We could talk to one another long before we developed written language, but there are prehistoric paintings. They appear to be images of things people are likely to have seen - it's a lot easier to draw a picture of an animal with distinguishing features for later reference than to sufficiently describe one. It's entirely possible that they are purely decorative too, which wouldn't really work with audio...

  3. Decoration. Writing is essentially a set of formalised pictures. Once you have pictures, it doesn't take a huge jump to get to a form of writing - think hieroglyphs as one example. And humans like decoration. Body painting was a very early discovery, and we're not the only ones who do it (although we might be the only ones who are purely aiming for artistic appeal - hard to tell!) with creatures like the Decorator Crab.

  4. Architecture. Once you get to trying to design more complex buildings, you want designs you can communicate once and refer back to. With audio-only, you'd have to either re-listen to the audio, or remeasure, when you got to, say, how wide the windows should be. If you can write it down on the plan, it's a lot easier. This gets even more critical with bridges, for example.

It's entirely possible that the written language in this situation might be less extensive, but that would appear to restrict certain developments. For example, while audio computers are possible, there would be limits on speeds of data transmission. Think modems, but just sending really fast speech, which is decoded back to listenable speed...


Prior to the advent of audio recordings and the printing press (and, possibly, widespread writing), people already used such a system of recording and passing down knowledge. Oral Storytellers were a prominant position within tribes and cultures.

Similarly, prior to letters, writing consisted of pictures and symbols.

Neither of these prevented a written language from being developed.

Writing is such a powerful tool for storage and recall. When you add the printing press, there is nothing like it for the mass distribution of information until electronic delivery.

Publishing millions of copies of "Common Sense" could not have happened with parrots, or wax, or storytellers, or even prior to the press.

To recap, the written word would not be prevented in such a world, and the traits of such writing would be that of what we have.



The invention of written symbols for numbers and lists of who owes what to whom occurred earlier than 'proper' writing

Handing the king a some clay tablets or a few scrolls with details of all the supplies his army needs or the tax his barons owe, might be easier or more convenient than giving him a flock of parrots. And as mentioned, in Matthew's answer, perhaps the king wants to keep this information secret, rather then have people overhear it.

Also in the modern world, paper and electronic tickets, permits, passports, work identity badges, shop receipts, bank statements and the like would be faster and more convenient to carry about and read, than to listen to a recording.


Civilisations invent currency so you don't have to move your wealth around in the form of cattle, bolts of cloth, and bushels of grain. Early coins have symbols on them to tell you who minted them, and later on coins also state what they are worth. So people will get used to symbols on currency.

You can see accurately further than you can hear accurately.

You don't want to have to be within shouting distance of that approaching band of horsemen before you find out if they are House Stark or House Lannister. Flags, banners and other heraldry will again be a form of symbols that humans use. It doesn't take much to start making scrolls/books of heraldry to teach your younglings what all these banners look like.

Meanwhile, you want signs of any kind from "Danger, black ice on road" to "Toilets are that way" to be distinguishable at a distance.

Noise pollution

You're at a railway station. The signs aren't written signs, they are parrots/recordings chattering away.

Platform 1. Platform 2. Stairs to platform 3 are over there. Escalators and platforms 4 to 6 are this way. Taxis and buses are that way. Fire exit. Staff only. Waiting room. Toilets. Ticket office. Arrivals are blah blah blah. Departures are blah blah blah.

It would be at best really annoying and at worst deafening.

Writing works in a power cut

Unless you are trying to read a book in the pitch dark, visual symbols and signs work when the battery to your ipod has gone flat, the mains power to the public address system has been cut off, and your parrot is grumpy or sick (or dead) and refuses to talk.


Yes, if they are humans regardless of any perfect audio recording, searching system, there always will be some form of VISUAL recording.

You seem to miss the point of WHY exactly writing developed and HOW.

Spoken language developed from animal sounds of warning, expressing distress, mating calls, etc. For the given purpose it served, it was perfect. It was instantly produced, could travel long distances and easily understood. It however had an huge limit, it relied on a single sense, hearing.

With increasing number of information to communicate, humans needed ways to better record and express the world around them. The world, perceived by their most used sense, their eyes. All written language originates from some form of symbols and pictures.

You can't communicate an information perceived by one sense through a medium which uses an other effectively and accurately. There will be a lot of miscommunication.

Describing a form, shape, etc. in spoken language will never be the same as showing it. Equations of all kinds, science and engineering in any form is impossible with our limited audio sense.

Closing thoughts: Even cavemen painted pictures of their hunts. Not for recording, but for better understanding.

A maybe better question is, would a race with much better audio sense than visual (sentient bats?) and the ability to record it, would develop a written language? That, I'm not so sure.

Predictions about the written language will be dependent on multiple factors: Society, Environment, Technology levels, History, Intelligence levels... how far from conception it is. The further in time it is, the less they will be like the pictures they are based on, and more of a mirror of their culture and history. Like Chinese characters... they are a fascinating study.


You must log in to answer this question.