Guide:Writing system
So you've decieded you want to make a script for your conlang. You probably want to do it because you've seen other languages have their writings or maybe you just want to be artistic, or maybe even your sister is reading your private diary so you want to be able to write it without her reading about your adventures! Well whatever the reason is it is quite the undertaking and you'll need to ask yourself a few questions.
Why do you want to make a script/writing system? What is the purpose?
This is important because it determines alot of the work that needs to be put in.
Is the answer you simply want to hide what you write so no one can read it? For sake of simplicity let's assume that whomever you are hiding it from has no knowledge in cryptography, a simple ceasar cipher might suffice then but you still want to be extra sure no one reads it by making your own symbols. Sure, replace every alphabetical symbol with one of your own and you're probably safe, but it is a cipher not really a script but it works! A hint is that you give common words like "the" their own symbol, it is the quickest way to crack your cipher to look for those simple words! You don't need to read much more in this article.
What? You're still reading? So you're not after a cipher but want a genuine writing system because other languages have it that does not match our writing? Then you have come to the right place! Keep on reading cause there is alot to learn.
Language structure
If you are doing a script for your own conlang or your own native tounge this one still matters. No no I am not talking about the long ass lists of conjugations, declensions and....wait what? you got more than just those!? Holy moly, well anyway that is not what I am talking about. The grammatical structure is the least important aspect of this but I am talking about the phonological and syllabic structure.
How many phonemes[*] does your language have? How are they arranged?
You probably ask now, "why would it matter?", it is simple, the type of script a language uses is affected by how it is structured. English which has a CCCVCCCC structure, 24 phonemes, 20 vowels or diphthongs gives the english language a total of 159 252 480, almost 160 MILLION possible syllables. So writing with a syllabic writing system or logographic like chinese and or japanese is clearly out of the picture, while most syllables are not used the number still grows to a point where you need to know tens of thousands possible symbols. In chinese you only need between 2 and 5 thousand symbols to be literate and thats hard enough!
Chinese on the other than has the structure of CVC, where the last C can only be m or n, even with 30 consonants and 30 tonal vowels (not real numbers just to illustrate) thats 900 base symbols sych a system would demand, a much more feasable amount to memorize. So what syllabic structure and the phoneme inventory does play a huge role.
But even that is not alone, grammatical structure does play a role aswell. Languages like chinese can avoid a more phoneticly based writing system because they technicly lack conjugations, declensions and the like directly on the word add words around to change meaning. So the words needs no more marking once they have been written out. A language like japanese which do conjugate their verbs is a whole other story. they MUST mark it in writing to give the right information. "But japanese uses Kanji which is chinese!" you probably say, yes they do. That is a result due to history but japanese have Hirigana and Katakana also, those are syllabic in structure, a much better fit for japanese syllable structure, to mark just phoneticly the endings verbs take.
Script Types
Thought it was over eh? OH NO!
You have already been presented somewhat to a few types but here I'll go into the writing types of the world and their implications and needs.
A feature most languages eventually tend toward is a phonetic representation but that can be divided into two categories, Syllable or Phonemic.
Syllable
These scrips writes out entire syllables, more or less, of the language in a distinct grapheme[*]. There are two of these.
Logographic
This script usually encompasses larger form of syllables, roughly around CVC structure and not larger. The script usually have a few hundred to several thousand symboles that must be memorized. It can be quite structured at first but like all systems the langauge changes and soon it is just a question of memorizing everything. This system fits languages with a smaller consonant inventory and simple syllable structure like the previously mentioned one. Chinese is an example.
- Grapheme count: Few hundred to thousands
- Phoneme count: Small
Syllabary
As the name suggest it goes for syllables but generally only CV form, if the langauge has such syllable structure (or close like only a single consonant can appear after) this system is quite useful. and important thing to remember with this system is that the grapheme for "pa", "pe" and "da" are all unrelated, you cannot see any connection between them. The phoneme inventory can be medium sized and it is quite feasable.
- Grapheme count: 30-100
- Phoneme count: medium
Phonemic
These writing systems catches the phonemes of the languages and generally fit any language, no matter syllable structure, that fullfill the criterias needed for them.
Abjad
An abjad catches only the consonants of a language. You might wonder how is that even useful? How would one tell apart words like "bet" "bat" "bit" if all you see is "bt"? The answer is simply that in european languages the vowels play a much more important role. In semitic languages the vowels role is lesser and consonants are more important, but even if it does play an important role you can still use it. "I bt 100 dllrs tht y cn't d tht", if you check that you might still be able to read what it says, context does everything!
- Grapheme count: 10-40
- Phoneme count: small to large
Abugida
Also known as Alphasyllaberies, the characteristics is that like a syllabery each symbol marks a CV cluster, but while "pa" and "pe" symbols are unrelated in syllaberies they are related graphicly to each other in an abugida. The vowel is changed by adding some tiny change, mark or whatnot. But it is important to notice that the original base symbol always have an inherent vowel to it that is modified and you need to add a symbol to even remove the vowel so it is a consonant alone. This language fits almost all languages but might not be suitible to languages that have too many non-CV syllables because otehrwise the vowel removal symbol will pop up quite frequently
- Grapheme count: 10-40
- Phoneme count: small to large
Alphabet
PHEW! this one we all know right? Yes and no, while western civilizations have grown around this there is still alot to it. It is capable to more or less represent all the basic sounds of a language it is by no means perfect or ideal. Vowels are more marked than an abjad for example but even within an alphabet the vowel marking is often at best just an approximation than something absolute.
- Grapheme count: 10-40
- Phoneme count: small to large
Direction
Everyone knows you read from left to right, up to down, as I call it LRUD. So thats natural isn't it?
Well no, you can do it in a total of 4 ways, LRUD, RLUD (Right to left, Up and downward), UDLR (up and downward first, then left to right, reading in columns rather than rows) and Lastely UDRL. You might think that it could possbibly be down and up, while it is a concievable thing no script system has ever utilized it.
Examples
- LRUD, Latin
- RLUD, Arabic or hebrew
- UDRL, Japanese
- UDLR, Mongolian
the direction in which it is written forms heavily how the symbols are written to minimize effort in that direction. so the direction does matter.