Published on September 19th, 2016 | by Luke Turpeinen2
Building Your Own World – Language & Names (part 2)
I Conlang And So Can You
Last week we went over some of the basics of language construction with the intent of building a naming language for your world building project. Languages that are made in this manner are neknown as constructed languages or “conlangs“. We went over an introduction to the International Phonetic Alphabet (IPA), the sounds those characters produce (phonetics) and how those sounds are used in language (called Phonology). We also briefly went over different ways to write down the sounds your language makes (known as Orthography).
This week I’ll take you through the process of actually putting together the information we learned last week, with an example language that I’ll make up right along with you. As you may know if you’ve read past world building articles, I have done some world building for a setting that I had given the working title “Kumarikan”. I knew that this title was problematic but it resonated so I kept it around as a place holder until I could come up with something else.
It was problematic in a couple of different ways:
1) It’s a sloppy bastardization of the name “Kumari Kandam”, itself the name of a lost-continent myth (like Atlantis). My setting doesn’t have much, if anything, to do with the myth itself so there is no reason to evoke it.
2) The name “Kumari” is an woman’s name in India, and surnames can start with “Kan” which means the setting would share at least part of a name with some percentage of Indian women (which, as you may know, is a large demographic).
3) The “Kumari Devi” is the Nepalese essence of divine femininity and “Kumari Puja” the practice of worshiping young girls as representations of said essence. In this sense, “kumari” means princess or virgin- which doesn’t intentionally describe my setting.
4) The last part of the name “Kumarikan” shares too much in common with the word “American” and now that I’ve heard it, I can’t un-hear it.
These reasons are all avoided by staying away from real world languages, especially ones whose associated cultures you don’t understand very well, if at all. Instead, make your own language so that you have a better grasp on its meaning, and a better control over the associations made to your words.
The first step in making your language is picking the “phones” (sounds) that your fictional language has in it. Letters are typically called “phenomes” and refer to the group of sounds that the letter form represents. This could be because of situations like English’s phenome l making both the /l/ and /L/ sounds, or something more subtle like aspirated/unaspirated stops.
In English, you push more air out of your mouth when you have a stop at the start of a word (consider the different p phones in the word “papa”). We don’t write those differences out, but some languages do and yours can too if you want. That’s why the words phone and phenome are import to understand.
I knew I wanted to make the language a little nasal, so I added the Spanish “ñ” to our three nasals (n, m, ng) and I’ll probably keep ng to the end of words. I kept stops/plosives as long as they are unvoiced- which gets rid of the phenomes b, d, and g. I got rid of most English fricatives, namely f, v, th, and h.
I kept s, sh, the normal English ch (which is really just a t + sh), and I added the ch sound from “Loch Ness” and “Chanukah” (it’s the Velar Fricative). I know I wanted w and y to be in the list, as well as r and both English l sounds.
English doesn’t have any “tap” sounds, but I like Spanish’s tap and trill (r and rr respectively). The trill is harder for some English speakers, but the tap is pretty easy (like Gandalf saying “Mordor” in the Lord of the Rings movies). I figure I can get away with only using one r phenome if the tap is only used at the start of words, like the English l sound division.
I also want to add the alveolar lateral fricative, which is a tough sound to make, especially for non-Welsh English speakers, but which I still like a lot. It is to the letter “l” what “sh” is to the letter “s”. A Navajo friend of mine taught me how to say it, as it’s used in Dine Bizaad and I’d love to include it. It’s also featured in Nahautl, which was the language of the Aztec empire.
Great! Now I know what consonant sounds I can use in my language, and I’ve already come up with a couple of rules for word-building that I’ll want to incorporate later on. I know that I have almost no retroflex or glottal sounds in my language. Next I need to come up with some vowel sounds.
For this language I chose a fairly simple list of vowel sounds. My thoughts were to go with basic vowels, like the 5 main vowels of Spanish (a, e, i, o, u) though I didn’t pick their exact sounds. Next I added the soft i from “swim”, as well as the soft u from “under”. I’m not going to add any dipthongs (vowel combinations), as I want to keep my vowels simpler than my consonants.
Now it’s time for some of the most simple of questions for your language: how do you let words form? The answer is the most important part of determining the feel of your language, perhaps only behind choosing the basic sounds available. What you’re really wanting to know here is what pattern of vowels and consonants are you going to allow to occur within your language’s syllables.
In writing your syllable structure out, the open letters are necessary, while parenthesized letters are optional. “V” means “any vowel phone” and “C” means “any consonant phone” while you can make up other variables to stand in for specific groups of phones.
Here I’ve listed out specific phenomes in lowercase instead of defining variables. Japanese is (C(y))V(n) while English is (s)(C)(r, l, w, y)V(C)(C)(C) and Hawaiian is (C)V(V). So Japanese can produce a one syllable word like “kyo”, English can produce “scratch” and Hawaiian “hui”.
Japanese sticks very closely to a CV structure typically, which is related to the writing system, and thus you get longer words that look like “katakana” and “Hiroshima”. In most languages vowels are going to be available as a distinct syllable, and will be the only non-optional value. Consonant sounds tend to occur in front of vowels, instead of after them (unless there are consonants both before and after the vowel). But as always, these choices are up to you.
As for my choice, I’ll stick with something relatively simple: (C(w))V(n, ng, m, s).
Now that I know what sounds my conlang makes and I know how they put those sounds together, I can figure out how I want to represent those sounds in writing. This is not exactly an easy task, and depending on your sounds you may have some tough decisions to make. Looking at my available phones I need to plan out which letters to assign to each sounds, which gives my conlang its own phenomes.
I personally like the idea of making art using this language later, so I think ultimately I’ll design my own type for the language but that can come later. I do want to see what the language would look like if they used the roman alphabet in-world, as I could also get double use out of this language as a futuristic or space-alien language in a sci-fi setting.
But if I ever want to use this language in a board or card game, I’m not going to be able to use a lot of funky letter forms and diacritics- I’ll have to constrain myself to the English alphabet. Though coming up with a version of the language that uses just one letter to represent each sound is really useful later when generating a lexicon.
Those are the consonants, with how I’d write the phonemes out in English, and how I’ll write them using only one character in a modern or futuristic setting. You could ask why the top transliteration isn’t sufficient by itself, and for the context of names within a larger product there isn’t anything insufficient.
But if you look at it closely, you can see it isn’t really internally consistent. There is no “g” letter out side of the combination “ng” and no “h” outside of the combination letters. The “ch” is especially funny, because neither letters occur on their own.
The technical transliteration makes more internal sense, but can’t be used as easily because the diacritics are by their nature foreign to literally everyone else in the world but you, right now. I really like the aesthetics of using ć as a slightly fricative version of English “y” alongside č as “sh”.
I learned Spanish from Chileans and Argentines who “shush” their “ll”s, but no one I know would ever be able to guess that pronunciation, so it’s best to do that sort of thing if you’re doing this mostly for yourself. Let’s look at the vowels now, they’re a bit more simple.
My vowel list is so simple that it’s pretty easy to just plug in the most English-like equivalent. The problem arises with what we call the “hard e” sound in English- as it’s also just the “i” sound in Spanish, but there is also an “e” and an “i” sound already represented.
We can’t use “y” because that’s already a consonant and I really don’t want to double dip across the lines like English does. So a double “ee” it is, doubled so as to make it “long” sounding. I’m not sure that “uh” will work for that sound all of the time, but if the vowel comes at the end of the word it should be fine.
For my in-world, technical orthography I’ll just use “y” for the “long e” sound because I used a “c” variant for the “y” consonant. That and I can just keep the lambda-like symbol for the “uh” sound. That gives me 23 letter forms: 16 consonants and 7 vowels. Now that I know how to form syllables, have letter forms for each phenome and I know how to write the language into English characters I can start generating words.
This in-depth article and our recent three article per week publishing schedule are thanks to the generous donations from our Patreon backers. If you wish to vote on the kind of content we produce, become a Patron today!