Comment by bradrn | Hacker News Reader

Comment by bradrn | Hacker News Reader

Parent Full thread

Not quite. Writing was independently invented about three or four times:

• Sumerian cuneiform was undisputedly the first writing system to be invented.

• Egyptian hieroglyphics was probably made with knowledge of cuneiform, but could have been an independent invention. All Western writing systems (Hebrew, Arabic, Greek, Latin etc.) are ultimately descended from hieroglyphics.

• All native writing systems of India, SE Asia and Oceania are descended from Brāhmī. I believe the current theory is that Brāhmī was descended from Aramaic script and hence also from hieroglyphics, but there are disputes.

• Chinese characters (hànzì) were an independent invention. Descended from them are Japanese hanja and kana, Vietnamese Chữ Nôm, Sawndip and some others.

• The writing systems of Mesoamerica (most notably that for Mayan) are of course a completely independent invention.

And then of course there are Hangeul, Cherokee, CAS etc., which were scripts consciously designed under the influence of other scripts. These tend to have rather unique structures compared to other writing systems of the world.

What we call alphabets, as distinguished from abigudas, abjads, and syllabaries, almost all derived directly or indirectly from Greek. Abigudas and abjads also almost all derived from Phoenician/Egyptian (from which Greek also derived). The independent systems of Mesopotamia, China, and Mesoamerica all ended up no further than syllabaries. What Indus Valley Script is isn't clear.

So while writing was indisputably multiply invented, the alphabet (i.e., representing all phonemes instead of syllables) itself appears to have been independently invented only once (excluding perhaps Hangul, which is somewhat closer to a rigorously-construct syllabary than an alphabet).

If you're unfamiliar with terminology:

* Alphabet = Every phoneme (consonant and vowel) has a distinct glyph. Examples include Latin, Greek, Cyrillic.

* Abiguda = Every consonant has a distinct glyph. Vowels are indicated by systematic modification of the glyph. Examples include Canadian Aboriginal Syllabics, Devanagari, Ethiopian.

* Abjad = Every consonant has a distinct glyph. Vowels are indicated by (optional) diacritic marks. Examples include Arabic, Hebrew.

* Syllabary = Every syllable (or, more often, Consonant-Vowel cluster) has a distinct glyph. Examples include Japanese kana, Mayan.

* Logograph = Every word (/morpheme) has a distinct glyph. Examples include Chinese, although Japanese and Mayan also include logographic components. Arguably modern English with emoji. ;-)

Hangul is definitely an alphabet.

The only reason one might consider it a syllabary according to these definitions is that the way the letters are arranged into a block form syllabic glyphs, but that’s more similar to writing direction (e.g. English left to right).

"Alphabet = Every phoneme (consonant and vowel) has a distinct glyph."

That's not really true of the Latin alphabet as used in English, is it?

"f" and "ph" could be pronounced the same, as in "fox" and "phone".

"c" and "k" could be pronounced the same, as in "cow" and "kiss".

So a multiple glyphs could represent the same phoneme.

Real languages, and especially phonetic drift over the centuries, do complicate the analysis somewhat. But the fact that some English phonemes in Latin script require the use of digraphs instead of single glyphs or there are multiple glyphs with the same phonemic value doesn't change the fact that the Latin script is providing glyphs for both consonants and vowels.

The same way Japanese is not a 'pure' syllabary (or moraic) script, as there are 1 mora syllables written with more than 1 symbol in an alphabet like fashion, such as しゃ (shi + small a = sha) or more recently ファ (fu + small a = fa)

There are alphabets not derived from Greek (but stil on the Phoenician/Egyptian family):

https://en.wikipedia.org/wiki/Mongolian_script

https://en.wikipedia.org/wiki/Old_Hungarian_script

https://en.wikipedia.org/wiki/Old_Turkic_script

> What we call alphabets, as distinguished from abigudas, abjads, and syllabaries, almost all derived directly or indirectly from Greek. Abigudas and abjads also almost all derived from Phoenician/Egyptian (from which Greek also derived). The independent systems of Mesopotamia, China, and Mesoamerica all ended up no further than syllabaries. What Indus Valley Script is isn't clear.

> So while writing was indisputably multiply invented, the alphabet (i.e., representing all phonemes instead of syllables) itself appears to have been independently invented only once (excluding perhaps Hangul, which is somewhat closer to a rigorously-construct syllabary than an alphabet).

Sure, I agree with this; I had thought that GP was using ‘alphabet’ in the informal sense of ‘any writing system’ rather than as a technical term.

No, I meant alphabet as opposed to writing. Sorry I should have been more clear.

I think the parent is maybe referring to something else, that I also heard in a documentary: that all writing systems with a "reduced" alphabet descend from the same source. I guess the japanese case is a counterexample to this, but the idea was, iirc, that writing systems appeared at different places, and followed this pattern:

* At first, people just draw representations of whatever they want to refer to.

* These evolve to symbols.

* Then, using the "rébus principle", you use the _sounds_ of each symbol to be able to represent things for which you had no symbols. This is the step where, say, the chinese and mayan systems are.

* Then, and this is the "reduced" part I'm referring to: you simply forget the original meaning of your symbol, and only take it for a basic sound. If I understand correctly, this happened when some people wanted to adapt hieroglyphs to their (spoken) language: they dropped the hieroglyph as symbolic representations, and only used them for their sound. The documentary I watched seemed to imply that all alphabets with this property (symbols have no meaning) descend from this moment.

It does sound a bit much. I may be totally misunderstanding it all, so any correction is more than welcome!

> Then, and this is the "reduced" part I'm referring to: you simply forget the original meaning of your symbol, and only take it for a basic sound.

This is usually known as the ‘acrophonic principle’, and indeed, all alphabetic scripts — and most of the other ones as well — are descended from Proto-Sinaitic, the first script to utilise the acrophonic principle.

Since you seem to know about this stuff:

* It seemed to me from that same documentary that pretty much all writing systems appeared independently from spoken language: In a sense, written and spoken appeared independently, and then the Rébus principle, and then this "acrophonic principle" made the writing system, and thus "language", sort of subdued to the spoken language. Is that true? It would intuitively seem to me that a written language is easier to organize into clean rules and principles, contrasted to a spoken language. Did "putting things in writing" help formalize such languages, or did it have no impact?

* Any good reference on these questions?

Thanks!

> It seemed to me from that same documentary that pretty much all writing systems appeared independently from spoken language: In a sense, written and spoken appeared independently, and then the Rébus principle, and then this "acrophonic principle" made the writing system, and thus "language", sort of subdued to the spoken language. Is that true?

I must admit to being confused as to what you’re asking here. What exactly do you mean by ‘appearing independently’, and writing systems being ‘subdued to the spoken language’?

> Did "putting things in writing" help formalize such languages, or did it have no impact?

Hmm… not sure. I suppose it depends on what you mean by ‘formalise’, and even then this isn’t really an area I’ve looked into. However, I wouldn’t be surprised to learn that the presence of a writing system assists in the preservation of archaic forms (not that this affects the rate of language change, mind you).

> Any good reference on these questions?

For writing systems in general, a good place to start is Omniglot (https://www.omniglot.com/), but I can’t think of anything more relevant to your specific questions.

>> It seemed to me from that same documentary that pretty much all writing systems appeared independently from spoken language: In a sense, written and spoken appeared independently, and then the Rébus principle, and then this "acrophonic principle" made the writing system, and thus "language", sort of subdued to the spoken language. Is that true?

>I must admit to being confused as to what you’re asking here. What exactly do you mean by ‘appearing independently’, and writing systems being ‘subdued to the spoken language’?

I meant that the writing system appeared at first, not as a way to "write down" spoken language, but as a language of its own (hence independent) with no relation to the already present spoken language. By "subdued", I meant that (again, from what I understood), the Rébus principle is the first "interaction" between the spoken and written language, and the "acrophonic principle" goes a step further in a sense, so that the written language is not anymore an independent system, but a symbolic representation of the spoken language (sounds -> symbols).

This is quite counter-intuitive to me, hence my asking about it.

> I meant that the writing system appeared at first, not as a way to "write down" spoken language, but as a language of its own (hence independent) with no relation to the already present spoken language.

This is incorrect, though an easy mistake to make. Though the symbols for words had no relation to pronunciation in the very first logographic scripts, written language was always connected to spoken language. At the very least, written language always utilised the same grammar as the spoken language: the written form of a spoken sentence was formed by taking each spoken word (or morpheme, if you prefer) and writing down the corresponding symbol(s) in turn.

Also, you seem to think that the rebus principle was a rather late invention. In fact, all known logographic writing systems utilise this principle. (You could even make the case that any writing system which does not use the rebus principle is not a ‘true’ writing system, in the sense that it cannot represent all words.) Even the earliest proto-cuneiform Sumerian texts utilised rebuses to some extent (https://oi.uchicago.edu/sites/oi.uchicago.edu/files/uploads/...):

> Phonetic writings generated via the rebus principle played a remarkably minor role in proto-cuneiform. More certain examples of phonetic writings include: the writing of the name of the moon god, Nanna, which is written URI₃+NA, where NA is the phonetic complement with the value na indicating that the graph URI₃ is to be pronounced nanna; PIRIG+NUNUZ, where the complement NUNUZ, has the value za, indicating that the composite graph has the phonetic value az(a); the aforementioned sign designating a reed, GI₄, pronounced gi, is used to express the homophonous verb gi “to return”; and the syllabic, that is phonetic, spellings of the city names Ša₃-bu and Gir₂-su (Englund 2009, pp. 9–10; Krebernik 2007, p. 43).

Alright, thanks for the clear up!

I'm confused about your quote though: is "remarkably minor" meant to mean "less important than expected"? Were there other forms of "phonetic writing" at play? If yes, then does that mean that there actually were case of symbols appearing directly as written translation of phonetic "concepts"?

> I'm confused about your quote though: is "remarkably minor" meant to mean "less important than expected"?

Yes.

> Were there other forms of "phonetic writing" at play?

At the time that quote was talking about? No, since at that time, proto-cuneiform (as they call it) would have been the only writing system in existence. Go forward a bit, though, and several forms of “phonetic writing” begin to appear frequently: rebuses (as in the quote), syllabic characters (e.g. [baʼugeš] was written BA-UG₇-GE, Edzard 2003), and characters for individual consonants (unattested in Sumerian, but Egyptian used them extensively).

> If yes, then does that mean that there actually were case of symbols appearing directly as written translation of phonetic "concepts"?

Exactly. (Even my earlier quote has examples of that.)

Thanks a lot!

Hanja is the Korean name for Chinese Characters (as used in the Korean language) https://en.wikipedia.org/wiki/Hanja Are you sure you did not mean "Japanese kanji and kana", or "Japanese hiragana and katakana"?

Also on the Chinese family are Khitan large and small script, and the Jurchen large and small script, where some of the characters seem to be phonograms grouped in syllabic blocks as in Korean Hangul!

Sorry, yes, I did indeed mean ‘kanji’, not ‘hanja’! (Pity I can’t edit anymore…)

> Also on the Chinese family are Khitan large and small script, and the Jurchen large and small script, where some of the characters seem to be phonograms grouped in syllabic blocks as in Korean Hangul!

Also the deeply weird Tangut (the only rGyalrongic language with an indigenous writing system). I didn’t know that about Jurchen, though; I’ll have to look into it.

Reply on news.ycombinator.com