Conlanging – Inventing Languages Step by Step (The Three Types of Conlangers)

This has been a long time coming. If I had a Euro every time someone asked me ‘How exactly do you invent a language, like, really?’, I’d have … fifty quid or so, because I really don’t talk to that many people … But hey, fifty quid! That’s something.

Anyways, you’re here because you have probably wondered the same thing. How does one go from A to B, or in this case from Nothing to A Whole New Language.

Short answer: You don’t. Long answer: You don’t, because you never start with nothing. There is always something, lurking in the deep dark pits of your soul that wants to break free.

But honestly, it’s true. Unless someone hires you to invent a language for their book or film or TV show, you usually have some sort of motivation, some sort of driving force that causes you to sit down and invent a language.

I have broken it down to three different types of conlangers. Chances are, you’re not one specific type, and your type might change from conlang to conlang, sometimes even while you’re working on it, but those three types are basically, or so I have found, a good guideline for you to determine how to proceed in inventing a language.

So let’s get started.

Step 1: Find Your Motivation

In order to invent a language you have to decide what its purpose is.

Type A: The World-Builder

You’re writing your book/screenplay/script/… and you have this entire new species or tribe or something. And now obviously you want them to sound differently. Because the reader/viewer/… needs to see that those people are not like the rest of us. So, how do you show that in the most convincing way? Obviously, you invent a language. Like any sane person would.

Type B: The Prompted

You might be sitting at home, reading a book, sipping your tea, or you might be out with your friends one day, shopping at the mall, when something strange happens. Out of the blue something triggers your imagination. It might be a word or sentence you have read, it might be something your friend said as a joke. And it was probably worded a little like this: ‘How weird would it be if …’ followed by some strange and linguistically rather improbable parameters that somehow stuck with you. So what, if you already know this isn’t how languages work and you will never be able to come up with something that actually works? Don’t tell me what I can do and can’t do.

Type C: The Scientist

You might have read papers about how ‘no language in the world does X’ or how ‘if a languages has Y it will always and without fault also have Z’ and decided that, hey, a language that breaks every single one of these rules (or universals, as we like to call them) would be a fun experiment. Or maybe you stumbled across some obscure hypothesis about language perception that makes you go ‘hm, let’s try that’. Or you might wonder what a language that was solely spoken by women would look like. Or maybe you just want to create a language that is the exact opposite of German or English … I could go on and on. Bottom line is, you invent a language just because you wonder ‘what if …’.
For science!


So, you see? Plenty of motivation for creating you own language. You never start with nothing. And once you figured out your motivation, it’s easy to determine the next step from there.

Step 2: Research

Scientists do it, authors do it, heck, even pupils do it (sometimes). So why do you think you can get away with not doing it?

Type A: The World-Builder

How do you research something that doesn’t exist? It might sound more difficult than it actually is – because, come to think of it, you already know a lot about your world and the people in it, don’t you? Even if it is nothing specific yet, nothing tangible, you have this sort of weird feeling about what is right and what is wrong. You know what I mean, right? For instance, when you talk to someone about your world and your people, and that person starts assuming things about them based on what you said and it is just plain wrong, but you can’t pinpoint the exact reason why it is wrong, you just have this weird notion that this is definitely not how your people operate. That feeling, is what I mean.

Now, onto the research part. Since you already have a general idea what goes and what doesn’t, all you need to do is write these things down. Write down everything you know about your people and the world they live in. Force yourself to go from the general idea to the specific word on the page. It doesn’t have to be stylistically flawless, just jot down quick notes, random notes, in no particular order, just the way they pop into your head. And if you end up with one single sheet of paper or fifty, it doesn’t matter. Write as much as you can, as much as you know, and eventually, you’ll come up with more and more details as the work progresses.

Type B: The Prompted

For this type, research is, in my humble opinion, pretty much impossible. At least, as far as I know. This weird idea that triggered you probably doesn’t even have any linguistic basis, so you can probably do whatever you want. The type of research for The Prompted depends highly on what kind of conlanging prompt got you going. The less scientific it gets, the less research you have to do – or so I guess. Personally, with my Prompted conlang, I don’t do any sort of research other than look up tones, stress patterns, nasal consonants, and all possible realisations of the vowel /a/. If you’re a Prompted, chances are, you already know what you need to research, depending on your prompt. Unfortunately, that is the most specific advice I can give without me knowing what exactly prompted you – but if you do want some more specific and personal advice, feel free to send me an email any time.

Type C: The Scientist

Chances are, this is not the first time you invent a language, and it’s even less likely that you’re a newbie to the field of linguistics. You probably study linguistics, or teach it to hapless students, and you have definitely done some scientific research before. I don’t need to tell you what to do. You probably know way more than I do.


Step 3: Start inventing (the phoneme inventory)

Okay, I feel like we’re about to enter territory where it becomes less and less easy for me to talk about conlanging without using linguistic terms. This is the ‘Advanced Conlanging Course’ after all, so at this point you might want to know a little about phonetics already. Because I’m not going to explain everything. We’d be here all week.

Onto the next step: I don’t think there is a right or wrong way to go about developing your conlang. Personally, however, I feel it’s easiest to start with the most basic: Sounds. Before you can create all kinds of words you need to know what your language sounds like, i.e. what kinds of phonemes occur in your language. IPA is your friend, and if you don’t know what that is, I highly recommend you learn it. Not only makes it things so much easier, it’s also more professional and allows other conlangers and linguists to read your language with ease.

So, grab your IPA chart and let’s get inventin’.

Type A: The World-Builder

Again, you already have a sense of what is right for your conlang and what is not. Same goes for the phoneme inventory. Start jotting down the sounds you already feel your conlang has and go from there. If you’re feeling inspired, you can add a few unusual phonemes as well, but beware of those you can’t pronounce yourself …
However, remember that this is your conlang, and you can go as crazy with your phoneme inventory as you want. Your conlang, your rules!

Type B: The Prompted

Chances are, you either have a specific set of phonemes that you have to include, because they were part of that prompt, or the prompt came from a different field of linguistics (e.g syntax) and at no point ever specified anything phonological. If your conlang falls into the first category, then it’s probably easier for you to create your phoneme inventory. If your conlang falls into the second category, well … just do whatever you want and you think sounds cool. Again, it’s a bit difficult to be general about something that can’t really be generalised.

Type C: The Scientist

You probably already have your colour-coordinated phoneme inventory printed out and pinned to your bedroom wall. I don’t need to tell you how to come up with one.

For those of you unsure what kind of phoneme inventory might be right for them, you can check out my other Conlanging posts about phoneme inventories. (Currently only available in German, however, but I’m working on translating them. In the meantime, don’t let Google Translate have a try. It won’t work.)

Conlanging – Das Lautinventar entwickeln (Teil 1: “Schön”)

Conlanging – Das Lautinventar entwickeln (Teil 2: “Exotisch”)

Conlanging – Das Lautinventar entwickeln (Teil 3: “Hart”)


Step 4: Phonetics and first words

To all my fantasy or sci-fi writers out there who weren’t prepared for things to go that scientific: Don’t panic! I promise, it won’t be that bad.
However, if you wouldn’t know what a lateral approximant is if it danced naked in front of your nose and you were only looking for a way to invent cool, meaningful names, I recommend my first article on conlanging, in which I focus specifically on how to come up with words and names for your fantasy world experience. Don’t worry, I got you.

And if you’re still here, great!

Let’s do some phonetics!

Now, you might say, ‘But I already have my phoneme inventory, can’t I go invent some words now, pleeeeease?’
Of course you can. In fact, delving deeper into phonetics and the phonetic properties of your language, and the invention of your very first words go hand in hand. How is that, you might wonder?

There are two things you need to know about your language (okay, there are, like, ten thousand more things, but only two that matter now). One: What does your language sound like? And two: How do you make it sound like that?
We have already covered the first point in Step 3, now we need to get to the point where we see how these cool sounds we have accumulated end up making even cooler sounding words.

If you’re new to this whole conlanging experience, it might be a bit overwhelming to come up with everything you need to do in order to actually invent a real language. I had to figure it out all on my own (mostly), so what I’m now passing onto you are things I have learned during my first conlanging invention. To you, the process of conlanging might be completely different, and again, there is no right or wrong way to go about it. I’m just here to give you an idea of how you might proceed.

What you need to figure out now, are the rules of how your phonemes fit together. You need to determine where in the word, where in the syllable each and every single one of your phonemes can occur. Can an /m/ occur next to a /k/? And if yes, in what specific environment? Can the onset of a syllable be or contain /mk/? Can the coda of a syllable be or contain /mk/? Can /mk/ only occur, when one syllable ends in /m/ and the following starts with /k/? How many consonants can occur in the onset of syllable, if any? How many consonants are allowed in the coda of a syllable, if any? Which phonemes can be the nucleus of a syllable? All this and more to consider.

Once you’ve determined that, you need to make rules for what happens when two phonemes that normally cannot go together end up next to each other. Those things are called assimilation rules. If you don’t know what that is, look it up. You’ll definitely need them.

So, where does this whole inventing new words come into? Answer: Exactly here. Because there is a difference between saying ‘/mks/ can only occur in the coda of a syllable’ and giving an example: ‘/mks/ can only occur in the coda of a syllable, e.g. namsk (= “bush”)’. It gives you something to work with and leaves you with a feeling of accomplishment.

And now you have created your very first word. Granted, ‘bush’ seems a bit anti-climactic, but hey, we’ve only just started.

Type A: The Word-Builder

The best thing about conlanging for a fantasy world is that you can really dive into the world you have created. That also means, you can create an actual proto-language for your conlang, with roots and sound shifts and everything for the ultimate conlang experience. As a matter of fact, I highly recommend doing so, because there might come a time when you describe an ancient artefact or something with an equally ancient inscription. Now, how likely would it be that this inscription is written in the modern version of your conlang? Not very likely. But if you create some sort of proto language, everything will be a lot more realistic – at least as realistic as fantasy will ever get.
Also, remember that languages have dialects. Your people might not all live in the same region. Grammar will differ. Not only regionally, but socially as well. Language is alive. Make it that way. And, again, because I cannot stress this enough: Break. The. Rules. You don’t need to explain why something happens. You’re not a Type C. You can just go about, doing what you want, and leave the analysing to the nerds.

Type B: The Prompted

Again, it’s slightly impossible to give you good advice here. If your prompt involved specific phonetic parameters, you need to incorporate them and work with those. Chances are, you don’t have a lot of freedom. On the other hand, if your prompt didn’t involve any phonetics, just do what you want.

Type C: The Scientist

I highly doubt you need any advice from me. Not that I could give you anything specific, because I don’t know what you’re working with – but what I feel I can say to you: There might come a time when you won’t find any remotely useful papers on certain specifications of your language. It’s quite possible that this is because this specific feature you want to include in your conlang doesn’t exist in any natural language (that we know of). I often found that once you start breaking linguistic rules you start noticing all the things that haven’t been researched yet. So, you might have to step out of your comfort zone and just bullshit your way through it. Every scientific theory is valid until proven otherwise … or something.

Step 5: Morphology

Ah, yes, morphology. Who doesn’t love inflections?

At this point, all I can give you are general ideas on how you might want to proceed in the creation of your language. Each and every single language is different from the next in so many ways, and there are even more ways you can go about inventing one, because you don’t need to follow existing rules, you can just make up your own.

The morphology part of your language will, without fault, make up the most of your entire notes. There’s a reason why there are so many chapters in a grammar book. Remember Latin? All those tables and conjunctions and everything you had to learn? Now you get to create that!

These are a few things you need to know about your language’s morphology:

Does it have nouns? Verbs? Adjectives? If yes, are they an open class, i.e. have a theoretically unlimited amount / can new verbs/adjectives/nouns be coined, like in English or German? If no, how does your language make up for that? What, if any, kinds of pronouns does your language have? Does it have nominal cases? Gender? How does your language mark possession? What about numbers and numerals? How does comparison in adjectives work? Do your verbs have tenses, modes, or aspects? If yes, which ones? How do you address one person? Two persons? Does your language have an imperative/hortative/benefactive? How are they expressed? What about congruency between verb phrase and noun phrase, and where is it marked (head vs. dependant vs. both)? Does your language have infixes, suffixes, prefixes, circumfixes?

Okay, raise your hand if your ears are ringing.

I know mine are.

Now, this might sound like A LOT. Why would anyone possibly need to come up with all of this??
You need to remember, you’re trying to invent an entire language here. You’re literally doing what usually takes hundreds of people over the course of millennia to develop. You might not think you need all of this but trust me. When you start writing poems in your language, you’ll quickly notice that you do need to come up with everything I just listed, and so much more.
Quick tip for each and everyone of you: Read about languages outside the European language tree. Look at linguistic papers on Arabic root morphology, Chinese tone, and ergativity. There are so many fantastic things out there that surely will inspire you.

Type A: The World-Builder

The more you want your characters to communicate in your language, the more grammar you need. That is a fact. Try translating existing poems or other texts into your language to figure out for which grammatical features you still don’t have an equivalent for. You’ll realise you can easily find another use for that inessive case you have lying around. Be creative with it. No-one can tell you what you can or can’t do, because – let’s face it – sometimes even natural languages do weird stuff linguists have a hard time figuring out.
What you always need to remember is: Natural languages don’t always follow the rules. If you have come up with a word or a morpheme that sounds incredibly bad-ass but you can’t use because the phonetic or morphologic parameters of your language don’t allow it – screw the rules. Use the word or morpheme anyway. It might be an import from another language, it might be a pidgin or creole, it just might be a remnant from some hundred years ago. Allow your conlang to break the rules. You’re your own boss.

Type B: The Prompted


I still don’t know what to tell you.
Basically, what I’ve told the World-Builder. Unless your prompt doesn’t allow it. Do you need to break rules? I don’t know. You don’t have to, I guess. Depends on what you’re trying to do …

Type C: The Scientist

Remember: It’s your language. If you have no bleeding explanation why something is happening in a specific way – just change it. If you can’t think of a rule to accurately predict the occurrence of Phenomenon A, then just cut it out. No-one will notice, because the language isn’t real.
Unless you need to have phenomenon A in your language, because it’s the entire reason you’ve started inventing it. Then go and talk to your fellow conlangers or linguists and ask them to have a look at the problem. Sometimes, a fresh perspective is all you need.


Step 6: Syntax (with a side of Morphology)

Don’t worry – we won’t draw any DP-trees here. Just keeping it simple. Mainly also because, truth be told, I’m not very good at syntax. It’s cool and I like it (most of the times, anyway), it’s just that there are things that I am better at. That’s fair. We’re all just human, you know. Life’s a peach.

I mean, if you want to draw detailed trees of your CP’s and DP’s and everything, please, be my guest. That is brilliant. Just don’t ask me to have a look and tell you what’s wrong when they don’t work out …

Now, the obvious question is: What do you need to know about the syntax of your conlang, if you’re not supposed to be drawing trees?

First of all, you obviously need to have the basic sentence structure: SVO, SOV, VOS, VSO, OVS, OSV, … and I think that’s all of them. Did I miss one? I feel like I missed one. Once you have your basic sentence structure, or BSS for your convenience, you also need to deterine what, if any, kind of subject, object or verb movement there is and when these movements occur.
Another very important point to consider: Does you language have a copula? If so, where and when and with which sentence structure does it appear?
Now, onto relative clauses: Does the relative clause have an internal or external head, does it follow or precede the main clause? What about adverbial clauses?
Very important also: Negation! How do you negate an entire sentence as opposed to single words or phrases?
And last, but not least: Questions. Polar-questions (a.k.a. Yes/No-questions) vs. open questions using interrogative pronouns (wh-questions).
All things you need to consider.

Also, the flexibility of your syntax is often closely related to your morphology. The more complex your morphology is, i.e. the more cases etc. you have, usually the more flexible your syntax can be. If you don’t have any way to distinguish subject from object, then your syntax will remain rather rigid to avoid confusion: ‘Alice sees Leah’ will only ever mean that Alice is the one seeing Leah, not the other way round, because of the distinctive lack of subject/object markers. However, if you did have markers, you might be able to switch it up: ‘Alice-Subject sees Leah-Object’ and ‘Leah-Object sees Alice-Subject’. Now, both sentences are perfectly understandable. Nominal cases do have a purpose, folks.

Type A: The World-Builder

Again, the fun part here is that you have what Type B and C (possibly) don’t have: An entire world full of other people and languages. So you can go crazy with your syntax. I mean it. Have fun and enjoy yourself. For example, my conlang Arassyani allows both SOV and VOS. Every time, all the time, without rules whatsoever. People can switch from SOV to VOS after conjunctions and the likes. In fact, that is quite common. So instead of saying ‘I went to the supermarket, and I saw the most attractive guy’ in Arassyani you would phrase it like this: ‘I went to the supermarket, and saw guy most attractive the I.’ Weird, but that’s language for you. As you can see, it’s fine to go overboard with syntax. Do what you would normally hate in natural languages.
And always remember: Other languages that exist in your fantasy universe will influence your conlang. One dialect might have a copula, another one might omit it under certain circumstances. In one dialect, SVO is the standard and OSV might be frowned upon, in another OSV is just as widely used as SVO. Be creative.

Type B: The Prompted

Do what your prompt asks you to and go crazy with the rest! Bonus: You don’t need to worry about dialects or proto-languages! That’s got to count for something, right?

Type C: The Scientist

I am 100% sure you know this stuff better than I do, so why don’t you maybe take over and help out the others (please)? Unless you’re busy with your own conlang, that’s fine …

Step 7: Lexicon

Last, but not least, the lexicon, a.k.a. your personal English-Conlang/Conlang-English dictionary.

Type A – C

I highly recommend you keep a list. Be it in Word, Excel, on paper or on your phone, I don’t care. Just write down every single new word you invent, including their translation, because you will forget it, you will forget where you wrote it down, you will not remember if you already have a word for this – and you will surely not sit down and learn every single word you invented until you know it by heart. You haven’t done it with Latin and French, why would you start now?
Pro-Tip: I mean, I know I said, use whatever you like to keep a vocabulary list – but honestly, the computer is the best place to have it. Tables that automatically arrange all your words alphabetically are a blessing to all conlangers alive! Plus, text search is your saviour when don’t remember if you already have a word for ‘bush’.
Ultimate Pro-Tip: If you happened to already have a word for ‘bush’ and now there’s two of them, don’t delete the second one. Keep it. It can be an import from another language. It can be a dialectal term. Maybe one of them is a taboo? Or maybe you use the first one when you talk to one person, and the second one when you talk to two or more people. Hey, those things happen.

Stay creative.

Further reading (articles are currently only available in German)
Conlanging: Erfinde deine eigene Sprache (Worldbuilding)
Conlanging – Checkliste
Conlanging – Den richtigen Ton treffen
Conlanging – Sprache schreiben
Conlanging – Das Schriftsystem entwickeln 
Worldbuilding – Dialekte und Sprachunterschiede



Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:

Du kommentierst mit deinem Abmelden /  Ändern )


Du kommentierst mit deinem Facebook-Konto. Abmelden /  Ändern )

Verbinde mit %s