Visit Page
Skip to content

Languages for Sale – Part 3

Posted in Just for Writers, and Language

This is the third of three posts on how to improve the use of languages in constructed worlds.  Part 1, an introduction of the topic for authors, is here. Part 2, an exploration for conlangers (the folks who invent CONstructed LANGuages) of what authors need is here. This final part is addressed to authors of fantasy and science fiction who might want to work with conlangers.

And if you want to see how this all worked out for my individual project, that report is here.

Why should my world building include language specialists?

A convincing world has history and context. It has artifacts from various cultures, some of the names of which came with the objects. It has transient fashions in names, and rulers or gods may be named differently from peasants. It may have non-human characters who don't use human phonemes to communicate.

Language also has history and context. It changes. It reflects the influence of other cultures. It memorializes conquest and trade. Each culture may have its own dialects and languages, possibly several. Characters from different cultures have different fluency in the default language (the one the book is written in).

Even if set in the future of our own quotidian world, the fashion in names will have changed, cultures will continue to mingle in unpredictable ways, new brands and technologies will come into existence and need names, and alien beings may make an appearance.

All of these things need names and convincing snippits of language to convey the appearance of a well-rounded historically-grounded plausibly realistic world.

Why can't I do it myself?

I can fake expertise (to some degree) in geology, biology, ecology, forensics, combat, medicine, physics, etc. And language. So can you. The question is: can we convince everyone?

We think nothing of asking experts to vet our descriptions of specialized knowledge, our plots dependent on mysterious Oriental poisons, our rendering of nuclear war, our development of killer robots, our construction of worlds that never existed. We accept that something that seems plausible to our superficial understanding in some area of expertise might actually not fool more knowledgeable readers, and we try to guard against that.

I don't believe that character said that, either
I don't believe that character said that, either

Language is both the same, and different. On the one hand, when we review our own work, we look for effective writing, for carrying the story along effectively, for investing our readers in our characters. That's language as a tool.

We can easily forget that there's another aspect of language — language as an immersive part of cultural and historical “reality” (however fictional). Think how we are thrown out of stories by anachronisms or by words put impossibly into the mouths of inappropriate speakers. Pulling words from some fantasy name generator does not give you a linguistically coherent set of proper names, and sprinkling in a few apostrophes won't fix the problem. Readers aren't fooled, even if they're not sure exactly what's wrong.

A few of these problems can be caught by editors, but editors only understand language-as-a-tool. They are not language experts (no, really). They don't understand historical linguistics, they can't speculate about how languages interact and evolve, and they can't detect the problems of verisimilitude in made-up languages and naming conventions.

That's what linguists do, especially that special group who invent languages in the conlang community. They develop languages and language families, create writing systems, and work out language evolution over time. Those of us in the fantasy and science fiction genres can benefit from their expertise.

What sort of services can I get?

I listed several things in the previous article, but let's take a look at them again from an author's point of view.


Lots and lots of names. Personal names, place names, language names, race names, and culture names.

The names within a language should look like they belong together. So should the place names, unless they come from an earlier population, in which case they have a different kind of integrity. (Think of all the various Indian place names in the US). If the names are descriptive (e.g., Red Wall for a mountain range), they should look like something that might translate that way.

How do characters from one culture or language make nicknames for other people on their team from other cultures?

Cultural & natural objects. Imported artifacts, food, religious practices, specialty curses, plants & animals. Brand names and technology names.

It doesn't take much, just a few bare words to give verisimilitude and the illusion of cultural interaction at the linguistic level.

Linguistic/historical/cultural relationships.

Historic depth. One language/culture may bear a clear relationship to another. Old versions may still be in use (the equivalent of Latin or Sanskrit). There may be high and low versions.

Linguistic breadth. For every race there may be a culture and for every culture there may be a language.

Characterization in dialogue. Foreigners speaking the default language (the one the book is written in) may use unusual grammatical forms or sentence structures reflecting their home languages. Lower class/upper class native speakers may also be distinguishable in a similar way.

Writing systems. Library & catalog forms. Materials & fragility.

Suggestive aesthetics. Subliminal grounding for readers. You can see that you might want to do this, but do you know how to achieve the effect?

Wombats may speak with low broad sounds.

Nomadic cultures may sound vaguely Arabic.

Ancient organized civilizations may vaguely suggest Chinese.

Pronunciation guides. For name indices in the books and to help with audio editions.


Dialogue, phrases, inscriptions, riddles, prophecies, clichés, sayings, jokes, insults. Poetic forms and poetry.


With the possible exception of external covers, interior covers, and chapter heads, this is an unlikely requirement for books (since it's so difficult to control images within text sections for ebooks).

How much work is it to construct a language?

Quite a bit. You need to define a phonemic system, decide how it operated in the past and then evolved, with sound shifts and simplifications and irregularities, into a contemporary form for the use of the book. Don't forget grammar (a complex subject, without which there are only single words) and rhetoric, the rhythm of sentences.

Slaving over a constructed language
Slaving over a constructed language

You need to define one or more writing systems over time, which impact the spoken language.

But you may not need all of that. It's possible to do a sketch of a language, sufficient to generate lists of names and some single words, which is all that most of us need most of the time. The deeper you go, the more work is involved for a conlanger.

You may also need a few key phrases, perhaps a prophecy or an inscription, maybe an alien marking, or a particular insult.

You may never need to know exactly how the language looks, its character set and fonts — that's really most useful for decorative elements in games and movies — but you'll probably want to know how the language is written down, what tools are used, how easily materials are preserved.

The clearer you can be about what you need for each language, when you describe it for a conlanger, the better they can set a price for the level of effort it will require.

Books vs Series

Working with a conlanger is a lot like working with a cover artist, but there are two very significant differences. First, if you lose your cover artist, you can always get a new one and, if necessary, reissue covers for the earlier books in a series to keep it consistent. Second, the requirement is more at the Series level than at the Book level. Once you've started naming characters and places and cultures and languages, you're stuck with them for an entire series, unless it's a stand-alone book. You may add cultures and characters, but you can't go back and change what's already there.

This has some implications. To begin with, the costs are front-loaded for a series. If you begin book 1 with members of 4 different cultures/languages on your team, well, you can see that a lot of the work has to be done at the start. The costs for later books would only be for new cultures/languages, additional namelists, and so forth. The good news is: once the work has been done for book 1, you might not need any more for the rest of the series, depending on how your plot develops. If you get a list of 50 names for Language A and only use 20 of them, the rest are still available for later books and stories in the series.

Another issue to remember is that a series has a longer working lifespan than a single book. Your relationship with the conlanger might not last that long (lives change). One of the work products you'll need from a conlanger is some form of hand-off language guide. If the conlang partner becomes unavailable, he should provide materials that would allow a different conlang partner to continue the supplemental work for the remainder of the series.

What does it cost?

In part 2 of this series I speculated about pricing from the conlanger perspective. In reality, this is a new concept and no one really knows what the prices should be. At this stage, conlangers have to give some thought about where their level of effort can best be applied and how much of their work can be re-used.

With a cover artist, you wouldn't expect the cover he produced for you of Warrior Babe in Leopard Skin to be used for any other author. On the other hand, you wouldn't care that he produced a modified Warrior Babe in a different composition in Tiger Skin for someone else, as long as they didn't look like direct or related copies and confused potential buyers.

Similarly, a conlanger could construct a full language and generate many namelists from it for different authors. The namelists that you receive, along with the name of the language and culture, should be exclusive to you, but if another author's book has different names and single words generated from the same conlang and the name of the language and culture are different, it would make no difference to you. Dothraki by any other name with different proper nouns would be effectively unrecognizable.

Consequently, you should expect exclusivity for the names and phrases you get from a conlanger, but not necessarily for the underlying language itself — that would be much more costly. The more of the actual language you want, the more you would need to think about exclusivity. In my opinion, exclusive full languages are only required for dialog use in media like games and film.

If the conlang community adopts this method of namelist production from a more limited set of languages or language sketches, the level of effort may be sufficiently reduced that prices can become quite reasonable. (Never free, like fantasy name generators, but you get what you pay for.) Be aware, however, that currently the recommendations of the Language Creation Society are that each author's ideal conlang requires unique cultural elements, and that each should therefore be unique and not reusable.

How do I find and interview a conlanger?

The Language Creation Society has a jobs board, and I imagine there are other conlang associations, too.

As you might imagine, conlangers have their eye on the jackpot — creating languages for movies, as David Peterson has done for Dothraki and Valyrian for George R R Martin's Game of Thrones. The notion that they could more realistically target the large potential audience of authors (mainly indies, but some traditionally-published authors, too) is new for them and it may take a while for the word to spread.

Contracts & Relationships

See part 2 of this series for details. In brief. I think these contracts should be modeled on contracts with cover artists, with a simple master agreement, and attached addenda for each individual project.

Remember, you're partners. Both authors and conlangers are in the same business — constructing imaginary worlds. A good imagination is a requirement for both of you.



Subscribe to My Newsletter

...and receive a free ebook: The Call, a short story that precedes the start of The Hounds of Annwn.


  1. […] This is the first of three posts on how to improve the use of languages in constructed worlds. Part 2, directed to the constructed language community, is here, and part 3, guidance for authors in working with conlangers, is here. […]

    November 24, 2015
  2. […] This is the second of three posts on how to improve the use of languages in constructed worlds.  Part 1, an introduction of the topic for authors, is here. This part is addressed to folks who invent CONstructed LANGuages: conlangers. The third part of the series, which provides guidance for authors working with conlangers, is here. […]

    November 24, 2015
  3. >> currently the recommendations of the Language Creation Society are
    >> that each author’s ideal conlang requires unique cultural elements,
    >> and that each should therefore be unique and not reusable.
    As an experienced conlanger I totally have to agree with the position of the LCS here. Language and culture are intricately connected in many ways, and this tends to show up especially in those parts of a language most likely to be needed for a novel, for instance in the meaning of personal names, in the metaphors used to describe religious beliefs and technological or magical artifacts, and in conventionalized expressions that are prone to be quoted verbatim for cultural flavor. Consider e.g. the Dothraki greeting “hash yer dothrae chek?” (“how are you?”, literally “do you ride well?” – perfectly suitable for a culture of nomaidc horse-riders) or as a natural language example the fact that English “goodbye” derives from a contraction of “God be with you!” In order to be able to carry convincing metaphors, a constructed language must be tailored specifically to the culture of its speakers, and thus be a unique creation. This takes significant effort, but it’s worth it!

    November 26, 2015
    • My issue is the “one size fits all” aspect of the LCS position. (Or two sizes, to be fair: language sketch from scratch or full language).

      If an already existing faux-Arabic, say, existed, with a list of cultural components, and it fit my basic cultural requirements for a fantasy world language that suggested those components due to its suggestibility of real-world Arabic, then I would be perfectly happy with a re-use purchase (with the serial numbers filed off, e.g., language/culture names, unique names/vocabulary). Maybe I could see a list of different faux-Arabic samples, from different conlangers, with different cultural details, and take my pick.

      Certainly there is a middle ground where from-scratch languages even at the sketch level for names would be required (and I’ve signed up for that myself), and a full-language need for extensive use, dialog, alien first contact stories, and so forth.

      Nonetheless, I think the low-end of possible reusability is an unexploited area that might suit a number of potential buyers without creating an unfair paid work effort for the conlanger. And we all know that the number of buyers goes up as the price goes down. The trick is in balancing price vs work effort, and the reusability suggestion is an attempt to do that for the low end of the market.

      It has the additional benefit of leverage, for the conlanger, something the LCS recommendations make impossible. Let’s say that a language sketch (from scratch) should be priced at a minimum of $150. Let’s say that a reusable language sketch should be priced at a minimum of $100. As a conlanger, it costs me $150 of work effort (simplifying the markup aspect) to make my faux-Arabic either way. But I can sell it again (maybe) for $100, which costs me very little additional work. And again. And again.

      Leverage is an important business decision for an artist. You can sell unique oil paintings, and you can also sell prints, engravings, and so forth. Nothing wrong with that.

      Will the market actually work out this way? I have no idea — depends on how many authors want expert assistance. The experience of freelance cover artists is that, yes, there are authors at the quality end who are willing to pay and, yes, there are aspects of creating cover art that are reusable.

      November 26, 2015

Leave a Reply

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