When I first set out to write my current project, I had an idea that involved at least one other culture. It was a key element that has always remained a part of my idea. Early on, I decided to focus on building the culture to make sure it was clearly foreign. Part of the reason for this is my terrible attempt at a sci-fi novel. One of the very many things wrong with that story is that it was from the perspective of an alien civilization, and it was not alien. There was no culture established and I hadn’t done enough to make them seem alien. When I started developing the story I have now, that stuck in my mind, so I set out to ensure that different people did have a different culture and voice. The starting point of this was, for me, language, as I have often heard that language defines a culture.

So, there it was, a starting point. To add to this, at around the same time my boss had suggested I read the Silmarillion by J.R.R. Tolkein. I had tried it before, but gave it another go, and fell in love with it. This launched me into a lot of reading about elvish as a language (languages really), and also the various writing systems developed for it. So, in developing my own language system, I looked at Elvish and Klingon, and also looked into oddities and elements of other languages. Currently, I have two con-langs that I use for naming places and also to add a bit of flavor into the text, though sparingly. Now, I still don’t have the grammar all the way nailed down for either, but I’m closer with Lotath than Swaerem. After all, these were intended for developing phrases, and naming places. It wasn’t so much intended for translation, although I find that to be a fun exercise. I would point to a vast array of different sources for all of this information, but it was all collected some 3 computers ago, and I no longer have most of the source material. In any case, it’s not a journal article or anything, so citations are more of a nicety than a requirement.

That said, here are some things I learned along the way.

Language really should define your culture. So, if you are creating language for a foreign culture, it should contain words that are ideas relevant to that culture. These words shouldn’t necessarily be directly translatable to English. If you find that you don’t have any of these sorts of words, it might be best not to bother with the effort of constructing a language.

It is my firm believe that constructed languages for an English audience should be English readable, or damn close. There are several guidelines I’d place around this one, but to pick on one for the purpose of illustration, is the use of the glottal stop. Avoid it as much as you can. Don’t eliminate it, just reduce it’s use. This IS present in English, but rarely -uh-oh is the most common American English occurrence. If not done well, this adds nothing to the foreignness of the language and makes it somewhat confusing to read. Not only that, this is one of THE MOST common things included in fantasy and Sci-fi, and often rendered as an apostrophe rather than hyphen or en-dash. If you insist, use it sparingly and not in names. That said, there is precedent for this, so you can get away with it if you want. I am largely (not completely) avoiding it though. The Dragon Riders of Pern uses the glottal in names: F’lar.

Another thing to consider is that English is odd when it comes to how vowels are written vs. how they’re actually pronounced. Before you get started, do a LOT OF RESEARCH ON VOWELS. If you take nothing away from this blog post, take that with you. Please. Also, define the vowels your language uses ahead of time and mostly stick to them. It would be a good idea to keep to those vowels familiar to English. I settled on the following set for Lotath:

ä – pronounced as ‘ah’
ë – you will find this in the English word bet,
é – You see this in the word resumé, though phonetically the same as the a found in rate.
i – As in the word bit
í – as in the double e found in beet.
o – pronounced the same as the letter o.
ú – pronounced the same as the double o in too.
ai – as in the i in mine.
oi – exactly the same as the oi in oil.

There are others I use, but these are the main ones, and I decided on them almost before I started building the language.

After looking at the list of vowels I chose as the primaries, you will notice a few things. First off. There is no ‘uh’ sound, or the schwa ‘ǝ’. This is the most common English vowel, and it doesn’t have its own character. It’s often given the u, but any letter will suffice (banana is often pronounced with schwa, for example). I concluded that because I couldn’t represent it in a manner that would be intuitive, I’d just avoid it. Western American English also often gets squishy on the short e and short i sounds, and so in cases where I think this will happen, I employ diacritic marks. I also use the diacritic mark when there would be an English tendency to actually drop the vowel sound, as in the ‘er’ found in diaper. There is some debate on whether or not the ‘er’ actually drops the vowel, but I’m in the camp that it’s phonetically not present.

Other things that make it impossible to read? Well, for starters, unless you’re writing a screenplay (if you are, ignore all of this – talk to a linguist instead), don’t try to use inflection, tone, or emphasis for meaning (If you are a screen writer, you SHOULD do this, because it would be awesome). I had started my first language like this, using an underline to indicate emphasis, but concluded it was not intuitive and dropped it. I also started thinking about a third language using a tonal system, but again had to conclude the English alphabet had no mechanism for representing this so it would be a waste of time.

The next bit of obvious advice is to avoid impossible stacks of letters like xmeqrdreg or zhredrgt. These just aren’t worth putting to paper. Perhaps it’s not that bad, but if it has no sound in English, just don’t bother, you’re wasting your effort. Every time the reader sees this it’s another opportunity to pause with the possibly of breaking the suspension of disbelief. On the opposite side of the coin, there are English sounds that are hard to get at because we don’t have a good alphabetic representation, like the ‘th’ in this, versus the ‘th’ in thing. Old English used Þ and ð to make this difference clear, but we don’t have those any more. You could use th and dh (as the voiced version of th), it’s what I do for my second language, but really it’s better to just not go there.

Pronunciation aside, be consistent. Select a handful of consonants that are preferred. It will help you build words that give your language a consistent feel. Another way to facilitate this is to use word endings or prefixes with some meaning. For example, you could choose ‘mes’ as short-hand for the word for county in your language, thus you could end up with Velmes and Nallames for two county names. Feels a lot more natural than Velamin and Nelmark. Not that the second choice wouldn’t work, but the first example provides the consistency you often see in a language. I might use the second in a case where I wasn’t trying to build a language.

Languages also often have a variety of features like case endings and declensions. There really isn’t any advice I could give to say what to use or not for this except that you need to look at these things and make a decision as to whether or not your language includes them – it probably should include some of this sort of thing.

The final consideration I’ll bring to your attention is that you are going to inadvertently break the rules just by not paying attention. This is perfect. Languages change and they adopt words from other sources. Having oddities in your language adds some authenticity that is hard to manufacture otherwise. If you screw up. Keep it.

So, after that less than complete discussion, research language before trying to assemble a constructed language. It’s a lot harder than just slamming letters together.