Homo sapiens have been around for 200,000 years, and spoken language is believed to have been around for 50,000 to 150,000 years.

Writing is a relatively new phenomenon. According to this source, writing is believed to have originated independently in multiple civilizations around 5,000 years ago.

If writing is a random invention that happens very rarely, this seems a really weird coincidence. Are there theories about why writing was developed around that time? Did the civilizations influence each other, or was the invention of writing facilitated by the development of specific tools or materials in that period of time?

  • 7
    Depends on how you define ‘writing’, I suppose. Australian aborigines have ‘written’ their stories on rock paintings for something like 40,000 years. Not fully-fledged writing systems, true, but an attempt nonetheless to recreate linguistic content in a permanent material. Other cave paintings and such things likely had similar functions, long before any of the recognised writing systems were even nascent. Commented Jun 5, 2021 at 22:04
  • 6
    The state of the society just call for some way to keep records. It is no wonder it is the time in the Bronze Age when agriculture was developed enough to support larger cities and structured society. Commented Jun 5, 2021 at 22:11
  • 2
    @JanusBahsJacquet thanks much for the comment. Perhaps I’ve been defining writing systems too narrowly. This actually makes a lot of sense: the Chinese characters were developed from drawings. Those were literally little cave drawings in the beginning, and only became an abstract writing system over time.
    – J Li
    Commented Jun 6, 2021 at 3:55
  • 2
    @VladimirF I see. So you are hypothesizing it is the need (and resources available from) agriculture that led to writing? That would make sense, as agricultural revolution happened roughly 10,000 years ago, which substantially narrowed the range of time. Bronze Age is even later.
    – J Li
    Commented Jun 6, 2021 at 3:58
  • 2
    More than just the agriculturul revolution. You need an advanced stratified society, not just "barbars" that did not need any writing in Europe even around 0 AD or even later. They could have easilly adopt the Latin ot Greek script, but did not need it. Commented Jun 6, 2021 at 8:08

4 Answers 4


From the source:

Full writing-systems appear to have been invented independently at least four times in human history: first in Mesopotamia (present-day Iraq) where cuneiform was used between 3400 and 3300 BC, and shortly afterwards in Egypt at around 3200 BC. By 1300 BC we have evidence of a fully operational writing system in late Shang-dynasty China. Sometime between 900 and 600 BC writing also appears in the cultures of Mesoamerica.

Only two of those are from "roughly the same period", Egypt and Mesopotamia. And those two are also much closer, geographically, than any of the others.

So the main theory I've seen is that proto-cuneiform, Egyptian hieroglyphs, and other early writing systems from that broad area (like Anatolian hieroglyphs) had an indirect influence on each other. While none of them were directly derived from each other, the idea of "writing can directly represent the sounds of spoken language" was likely transmitted through trade.

While direct evidence about the invention of these early writing systems is lacking, we see the same thing happening in America around 1820. The Cherokee syllabary is the result of an especially brilliant mind (Sequoyah) being introduced to the idea of recording the sounds of spoken language; he didn't know the details of any other writing system, just the fact that it was possible. It's easy to imagine a similar transfer of ideas between Mesopotamia and Egypt.

  • 1
    Note that the 1300 BC time frame for Chinese is when we find the first indisputably fully-fledged writing system, but it is clear from that system that a lot of changes and simplifications had already taken place. There is a fragmentary line of inscriptions going back a further 1,400 years or so that probably represent at least nascent writing: emblems of ownership or identification and such things. So for the purpose of this question, it’s not entirely unfair to take the Chinese evidence back to at least 2500 BCE, which is fairly close to the Mesopotamian/Egyptian evidence. Commented Jun 5, 2021 at 23:19
  • (Rolled back an edit that changed the first line; while the source is indeed from the British Library, that's not as important imo as the fact that it's the one source OP linked in their question.)
    – Draconis
    Commented Jun 6, 2021 at 5:23
  • 9
    @chepner The problem is, since it's undeciphered, we can't know if it's "true" writing (actually recording spoken language) or not. All the systems mentioned here were preceded by earlier recording systems that could handle accounting, represent different royals and gods, etc, but couldn't represent all of language in general, and unless we decipher the Indus script we can't know which sort it was.
    – Draconis
    Commented Jun 6, 2021 at 16:23
  • 2
    I don't believe the claim "he didn't know the details of any other writing system" is correct — you can see that Sequoyah's script has numerous letterforms that are clearly based on Roman print. The Wiki article on Sequoyah states that he used "the Bible as a reference along with adaptations from English, Greek, and Hebrew letters".
    – jogloran
    Commented Jun 7, 2021 at 3:57
  • 7
    @jogloran My understanding is that he had visual access to English, Greek, and Hebrew writing (and based some of his letters off them) but had not learned to actually read any of them. That's why his system ended up being a syllabary rather than an alphabet.
    – Draconis
    Commented Jun 7, 2021 at 3:59

I will answer from a different point of view. I will not care if it happened here at 3500 BC while at the other place at 1500 BC and consider it, as in your question, both as almost at the same time relative to the long history of Homo sapiens sapiens or even the long history of agriculture and larger settlements (Jericho 10000 BC).

People did not need make complex records before a complex stratified society emerged. Notice that even when people did use writing in a certain part of the world already, less developed societies nearby, which did have important trade connections to them, did not adopt it. Even much later. Germans and Celts outside of the empire did not (almost) need writing for a very long time. Slavs did not need writing before creating better organized states (or whichever word is more proper) that accepted Christianity. Typical neolitic (or earlier) cultures did not have enough complex and stratified society to develop writing.

It was the development of complex societies in Egypt, Mesopotamia, India, China or Mesoamerica that enabled the need for real writing instead of just simple "strokes and incisions" (Črьnorizьcь Chraбrъ - O pismenьchъ).

  • 4
    A society needs to also have lasted for a long enough time for us to find proof of they having developed writing. Imagine that in 10000 BC an agricultural city-state develops writing, but collapses less than 100 years later. And they didn't use clay or stone to write on, but more perishable materials. No matter how amazingly well developed their writing would have been, nothing survived, so we don't know about it. We know about later civilizations, because they were more stable, lasted longer, therefore the chance of something surviving until modern times is much higher.
    – vsz
    Commented Jun 7, 2021 at 12:32
  • Or alternatively, we might only know about writing systems from these civilizations because they are the most recent. Even if there were a number of equally-stable cultures with well-developed writing, say 50k years ago, they might not have used very durable material for writing (the choice is after all is a mostly cultural). Material is both re-used and degraded and our archeological coverage is far from comprehensive. Its far from trivial to experimentally distinguish both scenarios. @vsz
    – user0
    Commented Jun 7, 2021 at 18:37
  • This is like someone (I think Noam Chomsky) said, about research on teaching chimpanzees to speak, it seems they probably could if they wanted, but there's nothing they need to say.
    – RedSonja
    Commented Jun 8, 2021 at 8:48
  • In what language is the final parenthetical note (Črьnorizьcь Chraбrъ - O pismenьchъ), and does it mean exactly "strokes and incisions", or something else? I recognize the Slavic root "pis" (write), but not the rest.
    – Alan
    Commented Jun 8, 2021 at 14:04
  • 1
    @Alan It is Church Slavonic, "About letters", the title of the text. "using strokes and incisions" is "črъtami i rězami". Črnorizьcь Chraбrъ is a title and name. en.wikipedia.org/wiki/Chernorizets_Hrabar Commented Jun 19, 2021 at 19:48

In addition to what Vladimir said: Complex societies could only emerge when agricultural techniques were advanced enough to produce a surplus which could sustain a larger number of craftsmen, tradesmen and, tongue-in-cheek, unproductive managers like nobility and priests.

This surplus enabled the emergence of cities and palaces: Places that housed large numbers of people who were not directly concerned with producing food.

These had to be fed and sustained with food produced in the countryside. The food needed to be transported and stored in the palaces and cities.

Any large storage facility needs bookkeeping.

We know e.g. from indigenous Australians that the human mind is able to perform astonishing feats of memorization. The modern version of such a system are the German Ludolf brothers. Imagine a barn the size of half a soccer field, filled with piles of scrap metal parts 15 feet high:

The parts, stored in a warehouse, are sorted by a homegrown "cluster principle" with only Peter Ludolf knowing which part is in which pile. Each of the four brothers has a specific task. It's up to Peter to locate the desired part whether the customer is at the door or on the phone.

This could work for a while also in an god-emperor's palace. But it is not robust. People die, and memory is fallible. At some point people will start scratching pictures of the goods and a simple number system with lines on the doors and amphorae. From there it is a very slippery slope to a pictographic script.

We must assume that (indirect, relayed) communication and migration even over long distances like Mesopotamia and China existed, spreading ideas like agricultural technology and perhaps even the "bookkeeping meme". A culture that didn't improve their agriculture was bound to disappear in the long term; those that did at some point needed writing.


Adding a few bits to other answers and organising.

For writing to evolve:

  • A need has to arise
  • Need the means to write things down in a non-perishable way. (At least on the time-scale on which the information has value.)
  • Need a stable group where all can read/write/understand the system and hand the knowledge over to new people.

Most of these were realised around the time that agriculture developed:

  • Communities and then cities evolved.
  • Surplus means commerce is possible.
  • New professions evolve, creating additional types of goods.
  • One to one bartering is compounded by storage and vendors.
  • Ruling classes hoard resources for various reasons.
  • Agreements, contracts and longer distance commerce evolves.
  • A group can specialise in handling the above.

This subjective, but in my book, cave paintings and similar are not really writing systems as they are symbolic and only infer meaning. Plus they are from humans to some form of supernatural, so they are one-way and never really tested. They are projections of their wishes.

The bookkeeping/accounting phase on the other hand is meant to be a reproducible and unequivocal.

When it gets to language, you will find many opinions and arguments that of times end in subjective differences, so try to gather all of them and make up your mind.

One more note on the time comparisons: Human development, when it comes to culture and knowledge is close to being exponential, so when you compare times, unless some groups are close together to have frequent interaction, it makes sense to look at the timescales in a logarithmic scale. e.g. 3000 BCE vs 2500 BCE is not a big difference in this context. 1000 CE vs 1500 CE is a huge difference.

With the above in mind, developing agriculture created city societies around the same time. In Mesoamerica the process was similar, just delayed on the absolute timescale.

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.