Orthographic change over time

Question

This is related to a previous question of mine: Rate of linguistic change among geographically separated descendants of a common language

If you had 4 groups of humans with the same language and writing system, and these four groups were separated for approximately 15,000 years, how much would the writing system plausibly change in that period?

I think this question is worth asking separately as writing systems tend to change slower than spoken languages, and are often flexible enough to accommodate pretty significant phonological changes over time. Wholesale shifts to a brand-new writing system are exceedingly rare (Hangul script for Korean may be the only example). Beyond minor refinements in shape or simplifications, the major changes (such as adding or losing letters) seem to happen when one language adopts another language's writing system (see the adaptations made to the original Latin alphabet by other European languages).

In this situation however, each culture starts with the same language and writing system, and any developments are entirely internal. To rephrase the earlier question: even if over the period of separation the 4 cultures' spoken languages changed so much as to be completely unrecognizable phonologically or even morphologically, would it be plausible that their orthography would be broadly recognizable as having a common ancestor?

Answer 1

To the best of my knowledge there is no historical analogue for users of the same orthographic system becoming separated from each other. You're right to say change in writing is more "conservative" than change in speech, precisely because writing is how we have traditionally communicated across long distances. It's a prerequisite for writing to be a useful technology that it remains consistent across time and space, there's not really room for letters or symbols to be written differently, they can only be written correctly or incorrectly.

But 15,000 years is a long time, and replicative fade is a fact of life - no system can perfectly reproduce itself indefinitely, so, presuming all of your separated societies maintain a capacity for literacy over all this time, it probably is quite likely that their writing systems would be substantially different by the time the communities are re-introduced to one another.

But maintaining literacy for that long actually is quite a big "if". There are quite a few examples throughout history of societies abandoning literacy. We sort of take it for granted that it is natural and good for everyone to be literate, but if you're talking about a scenario where the level of technology has slid back to a pre-industrial state, where the vast majority of people are required to work on subsistence agriculture, there's really not much need for there to be much more than a small scribe class who can learn how to read and write. And then the smaller this scribe class is, the more vulnerable it is to being overthrown and abolished during times of turmoil. If some of your separated communities lose literacy and then re-invent it again from scratch at some later point in their history, all bets are off as to how different their new systems will be from one another.

Answer 2

I'll take your 15000 years and raise you 500.

Canterbury Tales is written in just over 500 year old middle English. Both the script and the words require significant study to understand.

https://nereg.lib.ms.us/wp-content/uploads/2020/02/chaucer.jpg

And that is with cultural continuity; no collapse, overthrow or anything else brutalized the English language from Middle English to Modern English. Just linguistic drift.

In 15000 years you could go through 30 such transformations, and probably worse ones. Only a handful of cities have been continuously inhabited over the last 5000 years, and all of them have been repeatedly conquered.

5000 years was long enough for most Indo-European languages to evolve from a single language, including Hindi. It again requires study to figure out how words are related.

While the existence of a written language could slow down mutation, it won't halt it.

I'd expect after 15000 years, it would require anthropology, archeology and linguistics to determine what the common language looked like, and people wouldn't be able to read each others texts. If a fragment of another branch's writing (a few 1000 words) was researched by a different branch, they would not be able to decode it, they'd need something like the Rosetta stone to have a hope.

Answer 3

"Would it be plausible that their orthography would be broadly recognizable as having a common ancestor?"

Sure, maybe, not unimaginable, if you know what to look for. For example, consider the following Russian words: сестра, брат, сын, вода, молоко, бук, язык, князь, месяц, and волк.

I have intentionally chosen words which are cognate with the corresponding English words, that is, they are descended from the same Indo-European words as their English counterparts. Five thousand years ago English and Russian were the same Proto-Indo-European.

And both English and Russian are written with scripts derived from the Greek alphabet, but at different stages of evolution of that alphabet; the Latin alphabet used by English comes from a western form of the Greek alphabet used about 2,500 years ago, the Cyrillic alphabet used by Russian comes from a form of the Greek alphabet used about 1,200 years ago.

So, the words are related (and their relationship is very much closer than the 15,000 years in the question), and the scripts are very closely related.

Do you recognize the words?

 

 

 

 

 

 

 

 

 

 

Not to hold anybody in suspense. The Romanized forms of the Russian words are sestra, brat, syn, voda, moloko, buk, yazyk, knyaz′, mesyats, and volk. The corresponding English cognates are sister, brother, son, water, milk, beech, tongue, king, month (or moon), and wolf.

Answer 4

It depends on your starting point. The usual assumption is that writing starts with pictograms and some numeral system, than quickly evolves to use some rebus scheme (the pictograms represent not the thing depicted but a word or syllable with a certain sound). Some writing systems essentially stay in this state.

The next step of evolution is a syllabary (the transition is easier for languages with simple phonotactics, e.g., Japanese created Hiragana and Katagana, but this creation is probably inspired by Indic scripts known through Buddhism). Many writing systems stop at this point.

The final step in the evolution is an alphabetic writing (in a broad sense, I include consonant-only writing system like Phoenician, Hebrew and Arabic under this umbrella term). Alphabetic writing is finally very stable what the general structure of the script is concerned, including a canonical alphabetic order of the symbols. It is unstable in the accidental aspects: Writing style changes all the time, compare Latin capital letters to Uncial style (Irish) or Fraktur, to the degree that the scripts become mutually unreadable; letters may be dropped or added to the alphabet depending on the needs of the written languages, the direction of writing may change, diacritics are added or dropped, to mention a few possible developments.

When you already started with an alphabetic writing at the time of split, chances are good that there are some similarities left even after 15,000 years. If you start with a huge script based on the rebus principle, it may be simplified to a syllabary or an alphabet at some point during the long period and the resulting writing will retain little similarity with the original, and even less with the independently evolving writing systems of the other populations. Without archaeological evidence it cannot be related to its ancestor and its sister systems.

Answer 5

As AlexP commented on the question, you are asking about a time interval that is three times as long as human history. So the short answer is "we don't know." In world building terms that means you can decide.

For writing to be stable there has to be old writing material that is both preserved and used as model for new.

Books don't really work, because they decay so quickly. Most of what we read is less than two decades old. And if you pick up a really old book and have problems reading it, you are more likely to think "what an old-fashioned writing" rather than "Oh no, we are doing it wrong now!"

Religious books are more stable, both because they are kept around for longer and because they are considered authorities, to some degree.

Still, if a book is copied carefully every hundred years it is still 150 layers of copies of copies over your time interval. Errors will build up.

However, monuments and buildings with writing on them has much longer time scale. They stick around for hundred or even thousands of years, proudly proclaiming that "EMPEROR KILLROY BUILT THIS TEMPLE TO THE GLORY OF HIS PAL JUPITER".

If somebody builds a new temple with inscriptions they don't want it to clash with the thousand year old temple next door. Here the number of copies of copies will only be 10-20, a much more manageable number.

To end this off, I would like to add counterexamples to the rule about no writing system ever radically changing.

The oldest writing we know was pictographic. At some point it changed to phonetic writing in most of the world. That was a very dramatic change. What they did was taking words with the right sounds and declaring that those pictograms should now stand for the sound rather than the meaning of the word.

Now, suppose a pictographic writing system was developed and spread to the entire world with minor changes.

Then phonetic writing was invented independently in several places. These alphabets are likely to be very different. First, you can chose different words to exemplify each sound, and second you can simplify those pictograms into letters in different ways.

Answer 6

Almost certainly not in reality but this is your built world.

Consider the slight transition from real Latin into what 'we' now call the Roman alphabet; the simple embellishments in French, German, etc; the more elaborate Central or Eastern European forms and the huge change into Cyrillic.

That variety evolved in a lot less than 2,000 years, during most of which time almost no-one could read or write and those who could were not 'scholastically' separated but generally had Arabic, Greek or Latin in common… if not more than one of those.

All known systems of writing evolved in a great deal less than 15,000 years. It's generally supposed they had no common root but how would anyone prove that?

Isn't that the other side of the same coin that says how few scholars less a century ago would have accepted the then-outlandish idea that all people descend from a single couple, however obvious it seems today?

Some experts suggest the very idea that even speech has existed for as much as 12,000 years is based purely on logic, not real evidence.

I stand to be corrected, and to me it seems the Question has no basis in reality.

Again, so what? This is your built world.