Kaltxì, ma frapo.
I hope you’ve all had an appropriately fun and spooky Halloween, and that November has started out well for you. Ulte ayngeyä tìmweypeyri irayo seiyi oe nìtxan.
I’ve finally replied to the comments I received on the last post—take a look when you can, as there are a couple of new vocabulary items there that you may find useful.
And here, for those who missed it, I’m finally posting the video of my Na’vigation talk from a few months back, which Tekre provided in a form that mercifully eliminated the technical glitches that were entirely my own fault. Irayo nìtxan, ma tsmuk!
The video is available on Google Drive here.
(If there’s any problem with the link, please let me know.)
As you’ll see, the talk was especially targeted to those of you who have created, or are thinking of creating, your own constructed language. The examples I gave, however, are largely from Na’vi.
One new thing in this talk was a brief discussion of orthography—that is, writing systems. So let me say a few things—or rather, ask you a few questions—about orthography as it relates to Na’vi.
For convenience, here are the relevant PowerPoint slides:
Some questions to think about:
A. Of the five types of orthographies mentioned—alphabets, abugidas, abjads, syllabaries, and logographic systems—which do you think works the best for Na’vi? Why?
B. In particular, what about a syllabic system for Na’vi? Syllabaries work very well for Japanese. Would they work for Na’vi too? (Hint: What is the significance of the number 8,690?)
C. The standard alphabetic system we use to write Na’vi is of course a modified form of the Roman alphabet. In most cases, there’s a one-to-one correspondence between sound and symbol: each distinctive sound is mapped onto one written symbol, and each written symbol is mapped onto one distinctive sound. But there are exceptions: the digraphs. These are pairs of letters that correspond to single consonants. The Na’vi digraphs are ng, ts, kx, px, and tx (and in Reef Na’vi, ch and sh).
For many, digraphs are unappealing—partly for logical or aesthetic reasons (“Why should a single sound need two letters?”) but also because of possible ambiguities. Think of English words like “mishap” and “cathouse” and you’ll see the problem.
Some of you may recall that early on, there were two proposed Roman alphabetic systems for Na’vi—the one we have now, which we can call the “popular” system, and one that eliminated some of the digraphs, which we can call the “scientific” system.
In the scientific system, the ts sound was written as c. Since c is not otherwise used in Na’vi, this is perfectly possible. And in fact some earth languages use c in exactly this way, so there’s ample precedent for this usage.
Also, the ng sound—the velar nasal—was written simply as g. Since g by itself is not used in Na’vi, this works well. And as with c, there is precedent in earth languages for this usage. (Pago Pago, the capital of American Samoa, is actually pronounced “Pango Pango.”)
All this being said, the scientific system was not successful and died out quickly. Why do you think that happened?
Hayalovay, ma eylan.
I apologize in advance for the long comment, I started writing down thoughts and they didn’t stop coming, so I ended up writing a small novel I fear hrh
Interesting questions! I have the feeling a syllable system would not work too well – the number I got is slightly different (the maths betrayed me I guess), but I assume the 8,690 is the number of possible Na’vi syllables. I honestly don’t know much about writing systems, but from my gut feeling, a syllabary would work better for a language with simpler syllable structure (e.g Japanese which it’s very simple CV syllable structure). If we’d have a different sign for each syllable, we could just have a logographic system instead and have less trouble with the current amount of words 😀 (although of course a syllabic system still would probably be easier to learn if it has some kind of regularity to it, which I assume it would, and it could be reduced to less characters + diacritics to show consonant clusters or something like that.)
I don’t really think a logographic system would work either for two reasons: In universe/linguistically, it makes sense to see this kind of system occur with isolating languages (e.g Chinese), because if there is no change happening to the words, words can be written the same way in all situations. I have honestly no idea how you would show infix usage in a logograph. pre- and suffixes might still work, but also be awkward in many siduations. If there is much morphology involved, I’d really not chose a logographic system. Out of universe I’d also say it’s probably too much work. I myself am working on a logographic system for one of my conlangs, and an alphabet (or any kind of other system) would be so much easier. I love logographic systems though, and if it weren’t for those two reason I’d definitely vote for one hrh
An alphabet would definitely work well. We have simple proof for that: It has worked for years in the community. Exchanging the latin alphabet with a new, in universe one would not change that. There is definitely tons of bias here as most people learning the Na’vi language are used to such a system, but that doesn’t change the fact that it is indeed something that works fine. If an alphabet were to be designed for Na’vi, it doesn’t specifically need to be 1to1 from latin to that alphabet for every word though – we have cases were phonological changes are represented in writing (layon -> layompin), but we have many others that are in fact not written when we write Na’vi. So looking at all of them and rethinking which would be represented in in-universe writing and which not might make sense. It also would be interesting to see how things that look rather awkward in current writing (thinking of stuff like lrrtok-susia, ta’lengna-eana syulang etc) are handled then. We handle them as special case – do the Na’vi do the same?
I hoenstly think a korean-style type of alphabet would be kinda cool!
For the abjad I made a small experiment. I took a few sentences and took out all the vowels to see if I could still read them. Of course I had picked the sentences myself and changed them so I knew what was written there. I tried to think of what other words could be there. Then I realized that the expirement is stupid and not needed because we literally have words only made of vowels, and the -ä case ending might disappear completely as well as , and ay-, and just imagine meoauniaea being reduced to mn. Just imagine mune and meoauniaea and man and mìn being all written the same. Yeah, I think that doesn’t work with a language that can have words that have no consonants.
Lastly, I definitely do not know enought about Abugidas to say much about it. From your description of it, I thought it might be problematic with having complex syllables with consonant clusters & coda, but looking for stuff online, it seems like there are abugidas for languages with such syllables.
Anyway, I think my favorite definitely would be a korean-style alphabet!
As for the popular vs. scientific systems, I think people are just used to see e.g ng for that sound. It is used that way in many indo-european languages. Same for ts. So for people that are not very interested in linguistics, that was just easier to learn. I think digraphs could also be part of an in-universe writing system. Maybe the Na’vi interpret px actually as another version of p? I think both, digraphs and completely different characters could be justified. It should just be made sure that the digraphs are not because of a bias coming from English, but for actual in-universe reasons.
That was my novel about writing systems without me actually knwoing anything about writing systems – I hope you had a nice Halloween!
Eywa ngahu!
I don’t really see the problem with transcribing synthetic morphology logographically. Even though Chinese is often said to be grammatically isolating, I think this description is a little misleading at least when it comes to modern Mandarin, which features not only abundant compound words but also a few suffixes, such as «-men» (the definite plural marker for animate nouns and pronouns), which is quite unproblematically represented by the logograph 們 (or 们 in the Simplified system), and even «-r», which, instead of constituting a syllable of its own, actually alters the pronunciation of the preceding syllable, represented as 儿. As for infixes, it is not too difficult to imagine them being transcribed in a non-linear way, as morphological alterations often are in Egyptian hieroglyphs for example, or indeed drawn physically inside other graphemes, which is not too unusual for logographic systems either.
With that being said, I don’t think it’s a coincidence that more grammatically synthetic languages like Egyptian, Mayan and Japanese ended up with robust systems of phonetic graphemes used alongside their logographs. I’d say it’s because creating a system where each grapheme represents one and only one morpheme in a grammatically synthetic language instead of recycling symbols for other things that sound similar would require a higher level of discipline, creativity and linguistic awareness than humans typically have. But would the Na’vi have the same psycholinguistic limitations if they were to invent writing?
Writing systems are really cool and interesting, but something I’ve learned in my work on Mando’a is: no matter what writing system you use, if the primary way people communicate in the language is via online chat, the Romanized alphabet will win out.
Mando’a has an (admittedly weak) alphabet system that people use often for decorative purposes such as tattoos or drawings/paintings/markings. Outside of the novel uses though, it has virtually zero appeal or application.
I fear Na’vi would suffer the same fate, no matter what in-universe writing system is used.
I’d say that each system has its advantage and disadvantages. If Na’vi would have a logographic system, like Chinese or Egyptian hieroglyphs, it would be surely interesting but this would mean that one would have to study 8690 possible symbols for Na’vi. Unfortunaltely, introducing such a system would increase the complexity and effort learning a language by having to learn not only vocab and grammar, but also the logographic symbols.
Regarding abjads, they may be easier compared to logographs, but they have their own issues that comes with using it. I have the strong feel that the abjad system will not work very well with Na’vi because Na’vi seems to be rather rich of vowels. As Tekre mentioned, how one will write “meoauniaea” or just “eo” or even “a” in a abjad system?
I am not sure about abugidas. They seem to have a similar issue like abjad mentioned above, but I am not sure if that system has the possibility to repressent single or clusters of vowels.
What also needs to be considered is how one will type these symbols? Even with Latin alphabet, we often saw people asking how to type “special” characters like ä or ì. This is of course seen from a practical point of view, and I must say, I’d love to see some kind of symbols or characters unique to Na’vi. (Just btw. in the first movie, one can see some “symbols” on Neytiri’s bow.)
As Alìmtsi mentioned, the big majority are using the Latin alphabet in online space (blogs, chats, etc.). But even the Latin alphabet is not perfect. Depending on the native language one uses, it has a significat influence on how people pronounce the characters that look indentical in a lot of difference languages. Like for example, the Na’vi ä [æ] is not the same sound as the German ä [ɛ]. This tragedy of the Latin alphabet cause a lot of pronounciation mistakes because one falls into that trap of indentical characters with actually different pronounciation.
On other hand, the Latin alphabet has a huge flexibility that other writing systems don’t have.
I am very familiar with dipgraphs as one who speak and use German that has a lot of digraphs that go up to heptagraphs, but also I am familiar with a language that has almost zero digraphs because it uses variants of the Latin alphabet like e.g. ț (a t with a diacritical comma) for the ts digraph/sound.
If digraphs are a real issue, then they can be replaced by a modified Latin character that is close the the sound it corresponds, or by adding new character symbols or even borrowing from other alphabets.
My guess of why the scientific writing for Na’vi died out quickly is,
– the majority of the learners do not assiciate the g character with a ng sound
– same also for c. I guess, a big part of the people are more familiar with the ts digraph that is supposed to represent the respective sound.
– I’d not be surprised if the above mentioned “tragedy of the Latin alphabet” also play a role here.
Well, I see I wrote a wall of text as well. Also sorry if some part may sound weird. I couldn’t wait and just started to type after I came back from night shift work.
It’s nice to think about how a Na’vi writing system would work!
Anyway, I figured I should let you know that there was a Learn Na’vi forum thread started in 2010 with the goal of gaining as many replies as possible. We realized that it is about 270 posts away from exceeding the maximum defined Na’vi number (32,767, or 0o77777 in octal). 😁 It keeps growing as we speak, and maybe in the future you’d read this after it exceeds that number.
Update: we did it
Irayo,ma Karyu Pawl.To be honest, I don’t have enough knowledge to answer these questions (but I am deeply impressed by them😊) and take part in related discussion😅.Actually, I have a question about ‘should have done’ in the counterfactual sense.
As we all know,the first situation is: ‘should’ refers to something that hasn’t yet happened.As in:
A. Sweylu txo po kivä. (‘He should go.’)
The second situation is:‘should’ refers to something that’s already happened.But according to whether s.o. did or not, it divides into two possibilities.
One is that s.o. did something and it was right thing to do.As in:
B. Sweylu fwa po kolä.(‘He should have gone(and he did).’)
Another is that s.o. didn’t do something but it was right thing to do.As in:
C. He should have gone(but he didn’t).
It is the question that puzzles me a lot.It seems that we still haven’t got a standard answer until now.But we recently discussed in the forum and got an answer.The sentence C can be explained as ‘If he had gone, it would have been best.’So, in Na’vi, it would be:
Zel sweylimvu zun po kimv(/ilv)ä.(‘He should have gone(but he didn’t)’)
Is it right?We’re looking forward to your reply.😊
I’m no scholar, but when I was learning Na’vi last year, I preferred the double-lettering like ng or ts, because it was easier to remember the pronunciation when learning and unless there’s a real need to change later on, I believe most will continue to use what they are more familiar with.
For example, I’ve been learning Japanese for 9 years now and when you first start you use romaji, but eventually you have to move on to hiragana and then a mix of hiragana, katakana, and kanji. Or else your learning will become stunted. Na’vi doesn’t require this transition.
Not to mention, while c and g aren’t used in Na’vi singularly they have their own distinct sounds to native English speakers and I’m sure others speakers, thus it may cause a hinderance to one’s thought process during pronunciation particularly when reading both verbally and mentally, especially without any (potentially) significant amount of practice.
I’d say a good middle ground would be to use alternative Latin/Latin-like characters that can take the place of these sounds. For example, I feel that the Eng character would be a great replacement for Ng, “Ŋ ŋ”, as its distinct and it looks the part of an Ng together. Of course Ng could be used by beginners or as shorthand with and English typeset, but long time learners and new ones would be able to quickly adapt this character into their writings. And to my limited knowledge the character is pronounced as proposed in all languages that use it, particularly in African and Sami languages.
(Some examples, I just looked up. I don’t know what they mean.)
ngeyä —> ŋeyä
ngopyu —> ŋopyu
ngungung —-> ŋuŋuŋ
For ts, I’m a bit partial to つ (tsu), because of my history with Japanese, but I know it’s distinctly out of place. Therefore some other alternatives I thought of were:
the now obsolete English letter, the long s, ſʃ
the Greek letters of Psi Ψ ψ (like the ps in lapse) or theta Θ θ (which is a t).
the Russian Cyrillic character of Цц or ts (same sound as Na’vi, I believe)
or the now obsolete Russian Cyrillic letter of Ѳ ѳ fita (was an f-ish t-ish, ts-ish sound)
There are some pros and cons to all. Long S is a bit odd looking and its lower and upper cases are virtually identical even in length. However, its modern relative obscurity makes it intriguing to new learners. Same with Fita. The Цц is a good fit because its sound is identical and make Na’vi easier for Slavic speakers to pick up. And Psi or Theta are a bit more familiar from fraternities and the like and their sounds are close enough to be a logic jump, however this just reinvigorates the original problem but for native Greek speakers. Which isn’t cool. That does improve the case of the thrown away letters of Long S and Fita as most don’t have preconception about their sounds as they are obsolete now. Also if you wanted to make the Russian Цц a bit more distinct you could use the old Novgorodian reversed ts, Ꙡ ꙡ. I know it look a bit more different than a simple mirroring, however that is just the font.
I didn’t mean to write all this and doubt anyone will read it, but hopefully the ramblings of a somewhat intoxicated 22 y/o will have at least some coherency.