Tuesday, November 19, 2019

An Idea For Useful Languages in RPGs

Languages are never fun in tabletop roleplaying games. People have come up with numerous ways to handle them, and their methods become especially absurd in games where languages are handled by leveled skill points ("Sorry, professor, but I can't learn Spanish until I go shank some goblins and steal their loot."). LotFP has a linguistic skill, for example, that you roll to determine whether or not you know a language; max the skill and you know everything. AD&D had "alignment languages" and people the first part of that term was even worse than the second. Some effort has been made to at least make the languages interesting. Meanwhile, most people just dispense with languages entirely. Here's a proposal.

Lexical Distance

In linguistics, there is a concept called "lexical distance" or "linguistic distance". It attempts to relate languages by proximity in order to demonstrate how much they differ. The details of how those determinations are made and the validity of the subject entirely is far beyond my knowledge. Instead, it helps break down a concept into discrete and gameable parts.

Lexical Distance Among the Languages of Europe

Among the Indo-European languages, we can see a memorable number of language families: Celtic, Germanic, Italic-Romance, Baltic, and Slavic. We also see Albanian and Hellenic families that have only one language. What's really convenient is that these 7 are precisely the number that most people can keep in working memory.

Some interesting things stand out. Greek and Albanian both have no nearby related languages. Are they particularly remote, or are they both recent empires who established widely understood lingua francas and widespread schooling (particularly writing) to ensure they don't drift? Per the chart, they are not widely spoken, but they might be Common, Trade, or Imperial in your game, something that everyone speaks and understand but never gets a chance to evolve.

Celtic languages are only spoken by a tiny number of people and the languages in the family diverge wildly. In the 21st century, they can be found in parts of Ireland, Wales, Scotland, and France. The great distances between these areas lead to significant divergence.

English appears among the Germanic languages but is probably where it's most apparent to many here the simplification going on. English monolingual speakers wouldn't really be able to speak with German monolingual speakers despite their shared family. English developed from Frisian languages spoken in the Netherlands, then diverged once it got across the Channel, then was significantly altered under French influence. But the point here isn't to accurately represent the history of language, it's to make language a mechanic that correctly balances simplicity and interactability.

Language Tree from Stand Still Stay Silent

With this tree diagram, we can also see exactly how languages developed organically as they slowly diverged. This chart is a lot more complex and so is less useful as a game concept, but it demonstrates a bit more of the history between them.

In our own history, the many branches of the Romance languages all derive from a former large empire and the divergence of its tongue. We also had large Germanic-speaking groups disperse widely and that goes hand in hand with the widely spaced branches seen in the tree diagram. Both developed into a multitude of languages, but the Romance stayed closer due to the cosmopolitan and expansive Roman Empire. But what happened to Romanian, which diverged early and appears somewhat alone among the Romantic languages? Named after but distantly related to its cousins, it was heavily influenced by the Baltic and Uralic while the rest tended to share amongst each other or with Germanic languages. This sort of thing can show up in your game too. Your notAtlantean diaspora peoples might all speak Atlantean-derived languages that diverged into different tongues in the various lands their speakers settled, like the DĂșnedain.

Speaking of, don't go full Tolkien. You just have to understand the relationship between languages, not invent them.

To use what I propose here, the DM will need to establish relationships between the languages in their game. They won't, and shouldn't, be as complex as the real-life ones above. But by understanding their common roots, you not only help implicitly detail the history of the setting but also help the players understand the relationship between their characters and the world.

Languages in the game

The problem with languages is that there is no good reason to include them and plenty of reasons to handwave them. As normally played, you either have the correct language and can talk to the NPC in question, or you don't have the language and that's it. And if the DM wants you to talk to them they speak your language. Reaction rolls may indicate they are friendly but I've seldom met friendly monsters who weren't also talkative in a language the PCs know. There is also never a way to be "better" with language that isn't a giant hassle defining highly granular levels of fluency.

What I've done here is add a stage between the binary known/unknown where the PC knows related languages. This is the level of normal "Languages Known". Now full fluency in the specific language grants advantage in social situations. It's not quite ranks of language fluency and it's not quite binary. Simple and interactable. A side benefit of this is that the PCs will likely end up with a variety of spoken language, as the game originally intended, and since there is a benefit for fluency it will lead to party spokesman being passed around.

Languages fall into several categories from a player perspective:

Unknown Language Players can describe what their PCs are doing to communicate. Nothing their PCs say can be understood. Time to play charades.

Known Family The next step up is having an idea of how the language works, some basic vocabulary, maybe a little grammar. Much of this will be similar between languages in a family. For example, fluency in Castillian, French, or Italian can all go a long way to understanding Catalan. The PC can freely communicate with anyone with whom they share a language family.

Known Language The PC has an understanding of the specifics of the language or dialect. They are fluent in the nuances, the conventions, and the colloquialisms. They achieve better results from any social interaction that involves talking. This also implies familiarity with the culture and customs of the primary speakers of the language in question. Whether that involves a DM judging the efficacy of an argument by intuition (much like judging whether a PC could lift something by referencing their strength), or a reaction roll to persuade, or (shudder) skill checks for bluff, persuade, etc. In the latter case, I guess you could say they have advantage, although that works a lot better if most of the people they meet don't speak the same language.

Lost Languages Like the texts in Death Frost Doom written in the forgotten language of a long-dead cult. Nobody speaks these, your PCs are never going to be fluent or need to be. But they may need to translate some text. They can use a rosetta stone (detailed below), but for longer works, Ben L. has written about nonmagical research.

Esoteric Languages Infernal, Enochian, Sidhe. They're never going to know these via character creation but can spend downtime to pick them up. Until they do they count as an Unknown Language and won't have a related Family to anything the players know.

Language Acquisition

At Character Creation

Richard Francis Burton
When a character is created, they gain whatever language is appropriate for their origin, including its family. Write both down to be clear. For fantasy Europe, write Romantic: French. Now it's clear that they have both the language and its family. With such a system, PCs should start with fewer languages in general, although scholar and diplomat type characters might know more. If they have several language choices (from high intelligence or class), treat languages and families as having to be bought separately, and family has to be bought before any language in that family.

After Character Creation

Characters can spend a downtime action to acquire a family, and a second downtime action to acquire a language in that family, as long as they have people around who speak the language in question and one the PC speaks. Double the time if they don't share a language. If more detail is required, tread a downtime action as a month.

If they travel frequently and far, they'll end up very multilingual, but this is pretty historical. Richard Burton, a British "explorer, geographer, translator, writer, soldier, orientalist, cartographer, ethnologist, spy, linguist, poet, fencer, and diplomat", who in his various travels in the 19th century picked up 29 languages. PCs can only wish to become so accomplished.

Literacy itself should probably be counted as a language in most time periods, after which PCs can read and write any language they know. For written text, literate PCs can carry a rosetta stone in their inventory. Without access to a rosetta stone, they either need to bargain with someone to provide a translation for them or head to a library to do some research.

Knowledge of language might be a reward from a magical patron. By saving the life of the Elf King, he might whisper in their ear the ability to speak Fey.

Rosetta Stone
An object that contains text in 2 different language families designed to aid in translation between them. Won't teach you the language but can aid in understanding its written form. Spend a turn to get a basic understanding of a text, meaning the DM should describe the text contents but don't read it to them. They can then spend an additional turn to receive the full text. If the text is an entire book, instead of turns, it takes equal numbers of downtime.

Source Unknown

For a game set in the Thirty Years' War

I initially composed this post solely about the lexical distance/language family concept, but then I remembered that I had written up some tables for generating characters for a LotFP campaign. They've been repurposed here in javascript to make example backgrounds for PCs or NPCs.

A couple of caveats:
  • Most of the participants spoke Germanic languages. This isn't weighted towards Germanic speakers but they still make up half of the total because they get broken down into the many different HRE states.
  • Because of the way I've defined language families and how languages do double duty as a cultural knowledge skill, it can lead to weird things like Transylvanian (which is just a Romanian dialect) appearing as a Romantic language equal to a full-blown language like French. 
  • Some of these people probably didn't show up at all in the war, but I included them because it makes it more interesting.
  • These aren't intended to be representative of the ethnic groups, languages, or religions of the people involved but instead provide an impression of those around at the time.
  • You can roll demon worshippers because of CKII.


  1. I think the jargon of profession, academia and religouns could count as their languages in this system, equivalent to languages known. As a real life example, my grandmother who's a physicist can communicate fairly fluently at English speaking conventions on the topic of science but would be hard pressed to hold a conversation in other contexts. Which could lead to interesting situations like, "we can talk to the dwarves, as long as we stay on the topic of metallurgy because one of the PCs is a blacksmith."

    Thieves cants, druidic tongues and other languages related to classes can also probably be folded into this system somehow.

    1. There's a thing in the Spire RPG called "domains". They way they use it is distinct to that game, but it covers the basic areas: Academia, Crime, Commerce, High Society, Low Society, Occult, Order, Religion and Technology. Shadowrun Returns does a similar thing with "Etiquettes," various classifications that unlock dialog choices. While I don't think you're going to get language family cross-compatibility, I think that you could distribute a domain (perhaps based on their randomly rolled background?) that works at the base level of language family (communication but not advantage) with anyone else that shares your domain.

      So your grandmother would have Academics (outside of a CoC game, I don't think I'd get more specific) and would get basic communication with any other Academic, but if she wanted to be persuasively fluent, she'd need to learn Germanic: English.

  2. I went on a tour of Europe with a couple of Uruguayan girls (it was a big tour group, not just me and them). They spoke Spanish of course, and found they could communicate fairly easily with French and Italian speakers. All Romance languages with enough similarities I guess. That always boiled my thoughts on language down to same language family = same language.

    1. Yup. If you know some French and Spanish, it's easier to get by visiting Barcelona because Catalan shares a lot of features of both. Similarly, while I couldn't ever communicate in any other Germanic language, I found that I could sorta understand menus and other basic text in places like Berlin and Oslo.

  3. Tolkien for all his invented languages only used them for history-stuff. Pretty much everyone spoke common in Middle Earth. Even the three groups of Orcs holding Merry and Pippin spoke common to each other in common.

  4. I like that you set "same language family" as the default, unmodified conversational assumption, and then hand out bonuses for "same exact language". As a game mechanic, that feels right to me.

    Fantasy Heartbreak Workshop and I were just chatting about this over on his blog too:

    1. Thank you, Anne. And I agree, thieves cant is a great use for Etiquettes, just like Gorinch's comment about academic lingo. The only thing I haven't figured is if it makes sense to keep alignment languages. Do they serve a real purpose?

    2. It's really difficult for me to imagine a world where there are individual species languages, and a common tongue, and a handful of occupational languages, and three "secret" languages that are each known by like a third of the population.

      I think they could almost make sense if there is no "common" language, and the alignment truly works like an alliance with one of three powerful "courts" of supernatural beings. (Imagine a tri-polar world where everyone is allied to Eurasia, Eastasia, and Oceania, for example.) In that case, alignment languages completely replace "common" within the campaign world.

    3. My one thought is that knowing the language is a prerequisite to learning clerical magic and they cannot simultaneously exist in the same brain. Their existence prevents the others from moving in and choosing to forget one is a requirement to learning the other. That way there are only 2 secret languages, chaotic and lawful, and if you can figure out that a person knows one you know something about them.

      If we wanted anyone but clerics to know alignment languages, there would need to be some benefit. Drawing from Moorcock, alignment languages might grant various boons like aligning with those same forces in the various Elric rpgs. Allegiance to chaos granted boosts to magic while law gave bonuses to skill checks in Stormbringer 5e, for example. They were rated as a percentile though, while this would be binary, so it couldn't be ported 1:1.

      Anyways, I don't think anyone really loves alignment languages so they might best be forgotten.