At last we come to the final part of the language creation process: creating a writing system. First of all, and most importantly, do you really need your writing system? Many languages around the world simply never developed writing as a technology. This is a perfectly valid option for your speakers if they're a preliterate society or otherwise wouldn't realistically discover writing technologies. Another equally viable option is that, rather than inventing writing themselves, your speakers borrowed a writing system from a different language. There are lots of fun spins you can pull on this concept, especially if the two languages have radically different phonemic inventories. But assuming you do want to create a script for this language that the speaker's devised themselves, there are a lot of things to consider before you can start creating symbols. Take a look at these four characters. Even if you have no idea what they mean, even if you've never seen them before, even if you have no idea what language they come from, you can still be confident that they all come from different languages. Why? They have different degrees of complexity, the line thickness differs, they have different levels of curvature, there are just too many inconsistencies between them. You would never ever expect all four of these symbols to occur in the same language together. All the characters that our language uses in its writing system should be different enough as to not be confused for each other, but still recognizably similar. There are a number of factors that go into creating this effect. Firstly, consider the medium: what instrument are your speakers writing with, and what are they writing on? Scripts will look radically different depending on these factors. Compare the straight lines and triangles of cuneiform, produced by pressing a triangular read into wet clay to the flowing lines of ancient Chinese calligraphy. Not only that but the instrument and medium of a script can change over time. The Latin script was most commonly carved into stone, resulting in a blocky angular symbols. But later when parchment became widespread and the script was written in ink, it was much easier to write with curves. For formalities sake, the older blockier forms were kept around for the beginning of sentences, Which is where we get our modern upper and lowercase letters. In German, they kept the capital letters at the beginning of every single noun, and in Hebrew, a similar stylistic decision resulted in a special form of certain characters that only occur when they are the last letter of a word. The second thing to consider is the type of system that the script is going to use. In alphabets, every sound is represented with its own symbol for both consonants and vowels, which might seem pretty intuitive and maybe even a little obvious to a speaker of a European language. However, abjads don't write characters for vowels only consonants, and they rely on the reader to infer what the vowels are purely from context. Syllabaries write every syllable as a separate symbol, typically with no visible similarity between related sounds, And abugidas or alpha-syllabaries write vowels as a modification of a base consonant symbol. Finally logographies encode words or concepts in symbols rather than sounds. Now, not all of these systems are going to be appropriate for all languages. Syllabaries and abugidas are best, though not exclusively used for languages with purely or mostly open syllables. Abjads are best for languages with a relatively small vowel inventory, since then the reader is more likely to infer the correct vowel, and Alphabets are best suited for languages with complex syllable structures. Also, the glyphs in a script will have a minimum level of required shape complexity, depending on the system. For example, in a logography, any given character will need to have enough detail to distinguish it from the potentially thousands of other characters in the script, and so will tend to be quite complex. Whereas in an alphabet, a character will only need to be able to be distinguished from a few dozen others at the most. And so can afford to be fairly rudimentary. But it's not enough to simply pick one of these systems and be done with it. None of the above systems were created out of the blue they evolved from earlier more primitive systems. In fact in the real world, writing has only been invented independently about five times, and every script in the world can be traced back to one of these progenitors. Even the Latin alphabet can be traced back to ancient Egyptian logographs. Just like we needed a proto grammar from which to evolve the modern grammar, We need a proto script from which to evolve the modern script. the logography is almost always going to be the earliest form of writing system to evolve as it's much more intuitive to Conceptualize drawing a picture of the information you're trying to convey than to create a separate symbol for every sound that you're saying. In other words, the idea of having symbols that represent individual sounds rather than basic meanings is a relatively new innovation. There's also a general trend for cultures that have recently developed writing to use symbols that are very complex and highly detailed, since they've only recently developed from straight-up pictures. Take a look at Egyptian hieroglyphs Can you imagine having to write out this symbol every time he wanted an M? Let alone carve it into stone? As such you'd only ever write something if it was really, really important and needed to be permanent, like carving the name of a deity into a temple altar or monument. This also means that generally only a certain subset of the culture's population will be able to read and write, usually those involved with the government and/or those wealthy enough to afford a lengthy education necessary to learn how to work with such a complex system, while the lower class may be entirely illiterate. This could even be used as a social tool to reinforce the divide between the upper and lower class, but that's a whole other topic. However, as writing gets more and more common, Especially when it gets employed for non-permanent use like in inter-personal letters or notes, the complexity of the symbols will get reduced over time until something like this ends up looking something like this. For your system, you should have an idea of the produce system it evolved from. What tools were used to produce it? You don't need to create an entire logography just for the proto-script, but it's good to have an idea of what the proto-glyphs look like and how they worked. In our sample language, we've said that our culture lives in a tropical forest. Perhaps they decide to carve their writing into trees. So if these glyphs are carved into wood then what constraints would that impose? First of all, it will be easiest to carve in the direction of the grain meaning it will likely be dominated by strong vertical lines. Carving into trees might also influence the scripts writing direction. A script's directionality may be fairly arbitrary. In Ancient Egyptian, hieroglyphs could be written right to left or left to right or sometimes even top to bottom. The semitic scripts like Arabic and Hebrew that descend from hieroglyphs settled on a consistent right-to-left direction. Whereas when the Greeks learned the alphabet from the Phoenicians, they went through a phase of what's called 'Boustrophedon' where in the writing direction alternates with every line. At some point, this fell out of favor and became replaced with playing left-to-right, which stuck around as the default writing direction for every subsequent European language. Since we're carving into trees, We could say that the dominant writing direction could be top to bottom, which would allow for the full height of a tree to be used for writing. Now that we have an idea of what the proto-system was like, we can start thinking about how it will change through history. Maybe our speakers eventually developed people like technologies using leaves, much like the palm leaf Tamil codices. Suddenly, the script is subjected to a completely new set of constraints. Because of the veined structure of leaves, they'll likely be torn by sharp changes in direction, so those strong straight vertical lines we've been using up to this point will start to get worn down. Also, since these leaf paper messages are most likely being used for short-term communication, this will impose a pressure for the glyphs to simplify, as they need to be written down much faster. And maybe to make things easier, characters start being used for the sound that they make rather than their meaning, a sort of sound-alike clue that in linguistics is called a rebus. For example, it's easy to represent basic concepts with logographs But how do you depict more abstract meanings? Like for example, this word 'tsani', 'to tell the truth'? Perhaps, rather than creating a separate symbol for it, our speakers instead find it easier to borrow the logograph for the similar sounding word 'tsame' or 'hand', so when a reader sees this symbol they'll have to determine from context whether it should be read for its sound or for its meaning. This rebus principle is ubiquitous across just about every logography in human history. Having a script that encodes just meanings and no phonetic information at all is actually vanishingly unlikely. But sometimes the readers of a given script may rely so much on the rebus characters that they begin to change the way that the script works entirely. For example, Maybe this character starts being used not just for the logograph for 'this' but also for the sound 'no', even when it occurs as part of another word, so a word like 'tlanapeno' or 'tea', could be written either with this logograph or with these symbols that independently mean 'two, me, this', Which obviously makes no sense, but stand for the sounds 'tlami', 'pe', 'no'. Now, these characters are being used to represent the sounds of individual syllables rather than entire words. And if this system becomes more and more standardized, suddenly we've got the beginnings of a syllabary on our hands. At this point, we need to consider exactly when the change from Logography to syllabary took place. Remember our time line of sound and grammar changes? We need to track changes in how our script works relative to all of those well. Our proto script was likely invented long before even the proto-language, So at what point did the rebus characters evolve into syllable characters? Well, remember syllabaries are best for languages with exclusively open syllables, which is exactly what our proto-language has, so perhaps the syllabary develops fairly early on before any sound changes led to increased syllable complexity. But once syllable characters are created and standardized, they might not completely out the logographs. Although the syllable characters are much easier to use and remember, it's still useful to have characters that encode meanings - in ancient Egyptian, Words were often written out with phonetic characters, but then immediately followed by an unpronounced logograph to remind the reader of the word's meaning. Perhaps our scripts might do something similar. So when we see these characters written together the first three served to convey pronunciation while the logograph isn't pronounced, It's just there to give you a hint as to what the meaning is as to eliminate any chance of ambiguity. So that'll serve us well for the proto language, But what happens when we come to our first sound change and we lose vowels between voiceless obstruents? Our previously CV language has become CVC, meaning we have some consonants that have no following vowels, so they can't be represented with syllable characters. How will this affect our script? The most likely result is that our speakers simply ignore the sound change and keep on spelling things the way that they're used to. This is called historical spelling, when the way a word is pronounced changes over time, But the way the word is spelled doesn't meaning the phonetic encoding of a script isn't 100% accurate. Spelling never fully keeps up with sound changes. Unless your culture discovered writing very recently in history, or are very very good at keeping up-to-date spelling, there will be at least some instances of historical spelling. And if the writing system is old the people are more conservative with their spelling reforms, there is likely to be a huge mismatch between speaking and spelling. How far does your language's spelling lag behind it sound changes? Take a look at our sound changes. If we say that the spelling system was last updated here, before the loss of word final vowels, Then there'll be a pretty strong correlation between words spelling and their pronunciations. On the other hand, if we imagine our spelling system hasn't been reformed since the syllabary was first developed then there'll be a huge dissonance between writing and speaking. One quick, but important note: all of this sound change versus spelling stuff does not apply to your romanization system. The only function of your romanization should be to transcribe the language as closely and intuitively as possible so that you or anyone reading your language knows for certain how it's pronounced. It's no use to anyone if your romanization system is just as weird as your language as spelling system. So we've said that the syllabary was first developed when the language had exclusively open syllables. For the first few sound changes our speakers probably won't mind continuing to spell things with the syllable characters, they'll just know that sometimes certain syllable characters are pronounced differently to how you'd expect. But once the language loses coda stops, we now have entire syllables in the written form that are no longer pronounced. Maybe our speakers decide to reform their spelling a little and get rid of these unpronounced characters just to be a bit more streamlined. However, this same sound change makes voice stops phonemic, meaning their distribution can't be predicted and therefore some words will differ only in whether a stop is voiced or not. Our speakers might want to find a way to represent this distinction. But instead of inventing brand new characters out of the blue, our speakers would probably find a way of modifying their existing characters. Perhaps they use the same characters for the syllables beginning with voiceless stops but mark them with a diacritic, or distinguishing mark, to indicate that they are voiced. This is almost exactly same as what's done in Japanese hiragana. These two changes (getting rid of unpronounced syllables and inventing the voicing diacritic) represent our languages first instance of spelling reform. This will work fine for a while, but by the time vowels are lost between nasals and obstruents, We now have loads of vowels that simply aren't pronounced. Perhaps this calls for another spelling reform, one where our speakers come up with a way to mark which vowels are not pronounced. Perhaps they could invent another diacritic to mark consonants that aren't followed by a vowel. A similar technique is used in many languages in the real world such as the brahmic script family. And what do we do about the [h] syllable once [h] gets lost from the language entirely? Well with our spelling reform, we might be tempted to get rid of them completely and replace them with the standard vowel characters, but be careful here. Notice how once [h] is lost and the preceding sound undergoes compensatory lengthening, the only clue in the writing system that these syllables are pronounced as anything different from a normal obstruent is the presence of a following [h] syllable. So keeping the [h] syllables around is actually quite useful. It's sort of like how English has loads of words that end in an /e/ that isn't pronounced, but we still write it because it lets us know to pronounce the preceding vowel differently. As the rest of our sound changes occur, they may not cause a substantial enough change to warrant any further spelling reform. Our speakers will simply get used to the fact that these syllables are palatalized when occurring before a vowel And that certain sounds simply aren't pronounced at the end of words. And that brings us all the way to the present day form of the language. For every word in your dictionary, refer back to how it was spelled in the proto-language and then carry it through the stages of your spelling reform to find out how it's spelled in the modern script. It may be a pain, but these kind of idiosyncrasies will help bring your language to life and make it feel all the more real. Lastly let's talk about punctuation. In the earliest form of many systems, there isn't any punctuation at all. As the language evolves punctuation may arise ex nihilo. A very likely form of punctuation is one that marks the end of sentences. Also quite likely as a method of indicating a pause or break in a sentence much like the comma in English. A language may also develop some way of separating individual words, most often with a space. In our case we can say that words started to be separated with spaces once the language switched to the syllable characters. Perhaps originally a single dot initially served as a break of any kind, but then doubling the dart emphasized a longer break such as at the end of a paragraph, which then became more widespread and became the generic mark for the end of a sentence. Many languages will be content to just have these basic roles filled but you can take it much further than that if you want to. And now, it's simply a matter of growing your language to your heart's content. Depending on your needs you could do well with just a few hundred words, or you could keep expanding it indefinitely. Well, that's about the full language creation process. Again, these are just the basic points that need consideration and there are an infinite number of details and variations for you to explore and pursue, but I hope this has provided you with some idea of how to get started with naturalistic conlanging. For any first-time conlangers out there, Let me know how you get on with your own languages and feel free to share any of your ideas or questions in the comments below. Until next time... [tsɔ.ˈpoɬ.wa pɛn.ˈti:m.ba] Thank you all for watching.
