On Inventing Languages

I love languages, they are something of a hobby for me. I have jokingly said to people in the past that I collect languages; which is true in a way. I’ve learned the basics of several, from Romance languages like French, Italian and Latin to Germanic languages like German and Norwegian, plus a few outliers like Welsh, Japanese and Klingon (yes, really – I even bought the audio tapes).

When I was a child, I used to construct languages of my own. They were very basic and lacked a proper grammar because they were essentially just complex ciphers of English but they were fun and I enjoyed spending hours on them while constructing imaginary societies and worlds that would use them. I even made a rudimentary Orcish language for Treasure Trap Live Action Roleplaying society; the basics of which were added to the society’s archives when I was its Archivist.

Everything you need for making your own language.
Everything you need for making your own language.

Now I’m building a more substantial setting for Sol Invictus as part of the development of a series bible to make a web TV show from, I’m looking into the language and society of 83rd century humanity. There are numerous languages and cultures within the Sol Invictus setting and I won’t be writing absolutely everything for absolutely every part of the setting but if I can put together something cool and substantial at the start, that’s a lot of ground work out of the way for future projects.

So I started looking up guides to constructing languages again. It’s been a while – the last time I read up on constructed languages was for an article I was writing for British Mensa’s Linguistics Special Interest Group (an article I never got around to finishing, now I think about it. Maybe I should dig that out and finish it off one of these years), so it’s been a while. I came across a brilliant site called The Language Construction Kit, which has some fantastic guides to making your own conlang. I thoroughly recommend it.

It’s also worth your while having a look at how David J Peterson (creator of the Dothraki language for HBO’s Game of Thrones) approaches the idea of past, present and future in his languages. Food for thought there on the idioms of speech.

Anyway, to cut a long story short, I’ve now set about building the first language for the Sol Invictus setting. It’s primarily based on Scandinavian languages with the idea being that English, German and French have merged into Swedish, Danish and Norwegian to create an amalgamated North Western European language. Then that amalgamated language has developed into the one being used in the Solar Republic from Sol Invictus.

I’ve called the language Ösklisk, as in the Sol Invictus setting it originated in a culture that developed out of late 33rd century Oslo. The Ö is pronounced essentially the same way it is pronounced in modern Swedish, as in the Sol Invictus universe Oslo is part of a territory where Swedish overtook Norwegian as the lingua franca a century or two earlier.

The first thing I set about deciding on was what effect this merging of languages would have had on the consonants are used by the people speaking the language. Thankfully the languages I’ve chosen to have been the primary influences on Ösklisk are reasonably similar in their consonant structure, as that made it easier to identify consonants that would not change a lot during the merge. I could have gone into researching consonant drift (like the way “d” words turned into “t” words during the evolution of Anglo-Saxon into English, for example) but this is my first real attempt to construct a proper language, so I’m keeping it simple where I can. I don’t want to bite off more than I can chew and then never get anywhere.

So where there was consensus between the languages, I kept the consonants they all agreed on. Where there were differences, I looked at where there was consensus between two languages and decided which sounded more like the language sounds I was hearing in my imagination. Where there was no consensus at all, I tended to default to Swedish because it fit the origin-in-Oslo theme. The result is pretty interesting, I think.









p b

t d

k g


f v

θ ð s z














Next up came a decision on what the alphabet would look like for a language like this. Yes, I know the next step should be to decide how the vowels will sound but I really wanted to play about with letters and numbers. It’s my language, dammit. I get to choose what comes next! Anyway, I have been thinking about how the Latin alphabet has changed from being entirely upper case in the Roman era through to the German obsession with capitalising nouns and all the way to the modern shopfront typefaces we see in England that have no capital letters at all.

I decided Ösklisk would have no capital letters (so the name of the language should be spelled “ösklisk”, I suppose) and that the alphabet should be sorted into letters that sound similar. I stuck to the Swedish alphabet because of the language’s origin in Oslo and set about rearranging it slightly. There are 29 letters in the Ösklisk language and they are arranged thusly:

a b c d e ä f g h i j k l m n ö å p q r s t o u v w x y z

I also started thinking about the problems that came with differentiating letters and numbers in my physics notes at university. Given that I’m inventing a language from the future, I decided that by the 83rd century we would have worked out a decent way of dealing with that particular problem. So numbers look slightly different in Ösklisk, too.

Zeros are now written with a diagonal slash through them, so they look like the Norwegian letter Ø. Ones always have a “cap”, to differentiate them from the letter “l” but they don’t have a horizontal line across the base. Twos are written in the French fashion, with a curl in the lower left, because it differentiates them from the letter “z”. Fours are always open top rather than coming to a point, because that differentiates them from the number nine. Sevens have a horizontal slash through their middles, to differentiate them from the number one.

These are all basic changes but they solve the problem of numbers looking like letters or other numbers so they serve their purpose.

There’s still a lot of work to be done before this language is anywhere near ready for building a dictionary but it’s been very good fun so far and as intellectual exercises go, I’ve been thoroughly enjoying it. When the language is nearer to being functional (i.e. once I’ve got a basic grammar down and produced some example phrases) I will release a PDF so anyone who is interested can take a look.

Leave a Reply

Your email address will not be published. Required fields are marked *