On an expedition to an inhabited planet, we encounter intelligent life. However, they are not as intelligent as we are. They have language, but it's less expressive and more concerned with the immediate - a strong vocabulary and simple grammar relating to hunting or foraging, say, but no deep abstract concepts yet.

I want to avoid just simplifying English. For instance, the "me-tarzan, you-jane" type language where most of the "primitiveness" comes from forgetting words. I want to make it sound like an actual language, in other words. In addition, I want to make it so that what it can express it can express naturally and fully. I go back to hunting, foraging, perhaps speaking about the immediate natural environment. Such a species would clearly be able to plan a hunt, for example, and would likely have a rich language for doing so.

What should I consider when making this up to achieve my goals?

EDIT: although there are many great answers, I think I should further specify to give more context. I'm not thinking of human-like hunter-gatherers. For instance, if we were to travel to the past, say, 50,000 years ago, and take a baby from that time and transplant them here, that baby would be able to learn our language and would not know that she is different in any way (time paradoxes aside). On the other hand, we can't do this with a chimp baby, although it is remarkable what we can do with chimps. I'm thinking somewhere between ape and human.

For instance, the language might only have one word for "danger" (why not? I don't need you to know what to run away from, just that you should run away) but I might have many words for "hunt" (depending on what I want to hunt, how I want to hunt it etc). Other than that, though, I'm sort of stumped as to what such a language would sound like.

EDIT 2: For the curious, I've done a bit of research and it seems that the most important feature of human language (compared to animal communication) is "displacement" - the ability to communicate about things that are not present, either in space or in time. I believe that, as a minimum, my creatures would need to be able to name a place and name a thing, so that, for example, the sentence "get berries at the hill" could be communicated, perhaps as simply as "berries hill". (Obviously this is in English to make this edit easier to write and understand). We could inflect the words to indicate meaning. For instance, "berries hill^" (raise last syllable, like asking a question) could indicate "we need to go get berries at the hill" while "^berries hill" could indicate actually asking a question (are there berries at the hill yet?)

a LOT more thought needs to go into this, obviously.

EDIT 3: I took a stab at outlining the language here: What are the limitations/expressive powers of my constructed proto-human like language?

  • 34
    $\begingroup$ I think one reason you might be stumped is that you're targeting probably the most difficult linguistic challenge imaginable: the razor fine line between having a language and not having one. We don't really know what's between apes and humans because we don't have many examples. You're really forging your own path in this region of language. It's always easy to talk about the extremes of "have a language" and "have no language." The middleground is the murky bit. $\endgroup$
    – Cort Ammon
    Apr 18, 2017 at 19:42
  • 3
    $\begingroup$ As a followup to @CortAmmon's comment, the good news is that you may actually come up with some truly new and therefore very interesting theories as you do your research. $\endgroup$ Apr 18, 2017 at 20:02
  • 5
    $\begingroup$ Mee-saw tinks dis is a good-ah question! $\endgroup$ Apr 18, 2017 at 20:03
  • 6
    $\begingroup$ Actual known protolanguages had anything but simple grammars. Ancient languages tended to have much much more morphology than English, and also unfamiliar morphology. Dual number? Middle voice? Aorist tense? Also, a very important aspect of human language is that it can convey messages (1) about unreal things (this very question being a good example -- the readers understand that the expedition and the planet we are speaking about do not actually exist) and (2) about the message itself. And using non-standard terminology won't help you in research (inflection != intonation). $\endgroup$
    – AlexP
    Apr 18, 2017 at 21:16
  • 6
    $\begingroup$ You don't have to be Tolkein or the director of the Game of Thrones TV series. Your task is not to construct a complete conlang. Your task is to advance a linguistic concept in a way that a reader can grasp it enough that still feels real by example. Like a TV studio, nothing that the camera doesn't see has to be spelled out or realistic. Focus on the concepts you want to highlight (very different from a human proto-language IMHO) and not on historical accuracy or completeness. $\endgroup$
    – ohwilleke
    Apr 18, 2017 at 22:05

14 Answers 14


A Little Goes A Long Way

Don't overtax your reader. By its very nature, your language will not have much of literary, poetic, or philosophical content. You want to give the reader a flavor without actually teaching them a new language.

Pick two to five words that have meanings that don't have good English approximates that really capture the essence of what you want to convey with your language and come up with specific, phonetically translated words to use in your text for those. Devote a lengthy scene or even a subplot to the process of discovering the meaning of one or two of those words, and a decent amount of coverage to the meaning of each other new word that is introduced.

A good example of how this is done in literary form would be Heinlein's Stranger in a Strange Land in which just a handful of words such as "Grok" are introduced but those words are used repeatedly and their deeper true meaning and fundamental foreign-ness as concepts is really explored.

A good example of how not to do this would be Frank Herbert's novel Dune which introduces so much foreign vocabulary so soon that it overwhelmed a lot of readers who weren't really determined. Ultimately, even that book had a fairly modest vocabulary set, but the book would have been much more accessible if the new vocabulary had been introduced at a more leisurely pace.

Not All Words Are Created Equal

Also, for these purposes, not all words are created equal. Introducing a new common noun that describes some kind of thing that doesn't exist in our world (e.g. the name of a snail that you suck into your nose to clear out your mucus) doesn't impose that much of a burden on a reader, and a proper name that isn't too hard to pronounce or remember imposes even less of a burden.

But, it is much more demanding on a reader to introduce words that convey more intangible concepts (e.g. a word that means hunger and pain and loneliness and fear or any of the above as context demands). So is a word that breaks the boundaries of our existing word categories (e.g. maybe instead of a verb for a particular kind of action like throwing something and an independent noun to describe what is thrown, there is a completely different word for every combination of an action like throwing and the thing that is thrown, a bit like irregular collective nouns in English such as a "murder of crows").

Show Don't Tell

Also, show don't tell.

Use Multiple Perspectives

Have a character try to communicate an idea that the more primitive language can't handle (like the future tense) express frustration over their repeated inability to get the idea across. But, perhaps using an eye of god narrator, portray in English an elaborate and cogent and fun conversation of some complexity within this simple language's boundaries between native speakers, to illustrate that it is not as simple minded as one might think.

Use Multiple Perspectives On The Same Conversation

Another interesting possibility comes to mind based upon an episode of This American Life on NPR about deaf people's cognitive experienced before and after learning sign language. Before learning sign language, they had a tough time remembering things and mostly expressed this when they did by re-enacting events that happened mime style. Only after they learned sign language did a more conventional memory emerge.

In your case I could imagine an outsider telling a story to someone with this primitive language that we hear from the teller's perspective with all its nuances, and then repeat the scene in a dream sequence of this native listener in a sequence that only repeats the greatly slimmed down version of the story that the listener was actually able to comprehend because the listener hadn't acquired all of the concepts that were communicated to the listener.

An Alien Language May Differ Not Just In How Ideas Are Communicated, But In What It Is That People Talk About

Another example of a real life proto-language would be dolphin communication which we are now starting to decipher. It turns out that dolphin communication involves a much higher proportion of proper nouns than human communication. It isn't that there aren't other words in dolphin language or that there aren't varied connotations of this basic core of what is said as in the "I am Groot" example in another answer. But, the core conversational content might be much smaller and might focus on two or three major subject areas like naming everyone and touching base with them in some way, identifying flora and fauna seen since the last discussion, and describing how to get from point A to point B, with not much discussion of other topics.

In the same vein, early written languages started out as a shorthand for keeping accounting records and wasn't intended as a complete reproduction over every idea or emotion that someone could have inside of you. You probably couldn't have written an emotional soliloquy about a teenagers feeling about their day in Linear A script or early Chinese characters any more than you could with the vocabulary in a set of baseball box scores. Maybe that is a kind of feeling expressed by these beings not with language but by touching and cuddling and stroking and crying and shrieking and eye rolling.

Pronunciation Through Font Effects

One good way to convey a novel kind of pronunciation is to convey it in writing with font effects. Perhaps some words are spoken with a throat voice and others are spoken with a nasal voice.

If the throat voice is conveyed with Courier type, and the first couple of times that throat voice is used that pronunciation method is identified, you visually convey the idea that non-English pronunciation is being used without actually having to transcribe the actually sounds made using the International Phonetic Alphabet (which may not even include sounds that aliens can make but humans cannot like sounds outside of our range of hearing or colors that we can't see with eyes built like ours).

Your ability, or a reader's ability to actually pronounce the words in the way that they would sound with this phonetic speech effect doesn't matter. The idea came come across even when the reader can't actually reproduce it.

Primitive Doesn't Mean Simple

Maybe you even add a feature normally considered "sophisticated" that is not present in English. Just because a language is overall simpler than English doesn't mean that it needs to be simpler in every single respect.

Navajo Source Of Information Tenses

For example, you could give this language the grammatical modes in Navajo that distinguish, for instance, between information obtained first hand and information that you know only through a hearsay account from someone else.

Topic Matter Noun Modes

Many African, Dravidian and Australian languages have different grammatical structures based upon the type of thing that you are discussing. There is one tense for talking about trees, another for talking about animals, and a third for talking about rocks.

Word Order

Another easy way to make a foreign language seem very different from English with minimal substantive difference or much of a learning curve is to follow the example of Yoda in Star Wars - use plain vanilla English but deviate from the SVO word order of English. There are even some studies that suggest that SOV is actually more "natural" if one is creating a language from scratch.

Complex Root Words Are Actually More Primitive

I would also deviate from other commenters regarding word complexity. Complex inflections and compound words may indeed reflect complexity and generalization and abstractification of a language. But, complex root words and many irregular forms can actually signal a more primitive languages.

Regularity and short efficient words are natural products of societies where people use lots of words on a regular basis with lots of variety in a society that is more seeped in culture in its essence. A less developed language would be less systematic, more idiosyncratic, and more clunky except for the most basic and universal concepts (mama, danger) the need to be expressed by even young children quickly.

A lot of complicated roots suggests that no one has analytically categorized words by type yet.

Don't Overdo It Grammatically

Whatever you do, however, you need to not stray too far from the concept that a little goes a long way.

Non-vocabulary language concepts are more of a burden on a reader than vocabulary based ones.

Two to five actual foreign words and one or two foreign grammatical concepts are more than enough. Translate everything else, adjusting point of view to do so if necessary.

  • $\begingroup$ +1 in general, but one nitpick. "Complex roots" in protolanguages? Proto-Indo-European had monosyllabic roots with a rather highly constrained form. Proto-Semitic had two letter and three-letter roots. $\endgroup$
    – AlexP
    Apr 18, 2017 at 21:37
  • 3
    $\begingroup$ @AlexP Proto-Indo-European like any human language was a complex and highly structured inflected human language of people with an increasingly complex economy whose large social scale made it possible for it to expand and conquer half the world. Also, the nature of the process of generating proto-languages in linguistics naturally favors this - it is inherently reductionist. Better examples might include Australian aboriginal mother-in-law languages or the words that children invent in early language acquisition which often has unreasonably long words (my first cat's name had eight syllables). $\endgroup$
    – ohwilleke
    Apr 18, 2017 at 21:45
  • 4
    $\begingroup$ @AlexP The core point is that no proto-language of any human language family is primitive. Indeed, proto-languages were by definition spoken by sophisticated societies that expanded while their neighbors did not mostly for reasons of greater cultural sophistication. The people who invented it had been refining their branch of the human language tree for 200,000 years or more and at least 50,000 years of diversification from an Out of Africa proto-language (or a few such languages). Proto-languages and primitive pre-human languages are two different concepts entirely. $\endgroup$
    – ohwilleke
    Apr 18, 2017 at 21:59
  • 2
    $\begingroup$ +1 as long and you dont smurf the smurf with that smurfing smurf, for Smurf's sake! Unles it's just played for smurfs, of course. $\endgroup$
    – xDaizu
    Apr 19, 2017 at 15:29
  • 1
    $\begingroup$ Really interesting answer - if you could add a few concrete citations or sources for reading more about certain topics, that would be fantastic! $\endgroup$ Apr 19, 2017 at 23:11

Make the language as dependent on the current context as you can. A more mature language will be capable of expressing ideas of things that are not related to what is around you, but an infant language is going to be focused more on the current reality.

I would focus on the pragmatics of your language over its syntax and semantics. Models such as Jakobson's 6 factors will help you hone your messages in this way. This lets you focus on what meanings actually get conveyed by your proto-language rather than focusing on what ideally would be conveyed.

Jakobson's model

Also, re-watch Guardians of the Galaxy, and see how much content is conveyed with "I am Groot!" Its actually pretty amazing what you can do without a complicated grammar and vocabulary.


  • 8
    $\begingroup$ I am Groot !! $\endgroup$
    – njzk2
    Apr 18, 2017 at 20:34
  • 3
    $\begingroup$ Hard to re-watch what you've never watched! $\endgroup$
    – Weckar E.
    Apr 19, 2017 at 9:54
  • $\begingroup$ @WeckarE. Pikachu, domo meep hodor, Timmy! $\endgroup$
    – Theraot
    Apr 19, 2017 at 14:16
  • $\begingroup$ @Theraot Well, I knew I shouldn't have clicked, and I did anyway. $\endgroup$ Apr 20, 2017 at 16:46

At the risk of sounding culturally insensitive, you may want to consider the way someone learning a new language speaks.

They often have little conjugation, as in only using the present tense.

Limited vocabulary, only words they've needed to get by thus far.

Few descriptive words. Think size, basic colors​, hot/cold.

For instance, I know I can ask for a bathroom and a dark beer in Spanish, but I'm far from fluent. Just enough to communicate basic needs.

The exception to this line of thought would be a very well developed sense of their surrounding world. Expect to hear very specific names for plants and animals that they see around them. Rather than general words like "tree" they would likely have names for every variety of tree, particularly ones that produced food.

  • 2
    $\begingroup$ Also when languages collide, things get taken from each in an interesting way. The word for food might be taken from the native language into the non-native one, while the word for a God might be taken from the non-native into the native. $\endgroup$ Apr 18, 2017 at 17:36

Sign language might be the answer.

Simple sign language can cover most daily activities, and limited simple abstract concepts are available (like emotion, love).

For primitive tribe, sign language is the most efficient and effective method. Spoken language are very limited because most communication can be done more swiftly and efficiently using sign language.


  • Distance - sign language can be seen from afar
  • Hunting - sign language can be used without alarming the prey
  • Wind - sign language is more reliable when used in an open, windy area
  • Instinct - basic sign language can be understood immediately without having to teach the recipient

Note: I'm talking about primitive sign language, not modern sign language

  • 1
    $\begingroup$ Reminds me Clan of the cave bear $\endgroup$
    – James K
    Apr 18, 2017 at 18:40
  • 6
    $\begingroup$ Actual signed languages are as fully expressive as anything spoken. What you're talking about is not true sign language. Perhaps it could be termed a gestural system? $\endgroup$
    – TRiG
    Apr 18, 2017 at 20:15
  • 1
    $\begingroup$ @TRiG actual, modern signed languages. There have been signed languages that were more basic: en.wikipedia.org/wiki/Plains_Indian_Sign_Language $\endgroup$
    – Jeutnarg
    Apr 18, 2017 at 20:56
  • $\begingroup$ @TRiG yes, I mean a primitive sign language when I wrote simple sign language. It will consists easy to understand gestures (like STOP, EAT, GO, COME-invite). Because they have not yet developed civilization, abstract concepts will be very limited. $\endgroup$
    – Vylix
    Apr 19, 2017 at 1:14

You will more convincingly convey an alien attitude with what your natives say, rather than how they say it. There are lots of people who come at things from other than a Western European / goal directed / lets do the deal sort of background. Talking to such people can be disconcerting if you come from the former background, especially if you do not expect it. It seems like they are purposefully avoiding the matter at hand. Maybe they are. Maybe they are shy, or do not want to be offend, or worry that they might jinx things by speaking out loud. Maybe in their world view they do consider the things they say directly relevant. Bring that out in the exchanges you write for your natives.

Here is an idea how.

  1. Reread Old Man and the Sea. The dialogue between characters is the very direct, in the moment simple language that you want for your people. It is powerful and it is not annoying.

  2. Your natives should talk like that. Your nonnative should not.

  3. Your natives should talk about things that seem irrelevant and irritating to your nonnative, who wants to get to the point. These things said by the natives could be anecdotes, or questions, or opinions. Use these to help convey the mindset of the natives. Possibly the natives never explicitly plan the hunt out loud, but convey to their comrades how it will go because of the comments and talk on other (seemingly random things). You could even leave that opaque and then have your nonnative marvel that anything at all got done, much less as efficiently as it did.

If you purposefully never have your natives address a concrete matter headon but rather tangentially or by allegory it will be a good rule of thumb for writing. Sort of like a person wanting a bribe might not ask outright for the bribe but will instead talk about lots of other things.

  • 1
    $\begingroup$ That's an interesting idea, but I'm not sure if it works in the context (see the edit above). I'd say, if anything, the natives would be straight to the point. No abstractions at all, wouldn't they? $\endgroup$ Apr 18, 2017 at 17:29
  • 3
    $\begingroup$ it depends. If the need was immediate (fire, danger, run) then yes, they should be to the point. What if there is no immediate need? Do they have stories they tell? What else do they use their language for? I like rewatching the Darmok episode of Star Trek TNG when I start thinking about how other peoples would speak in my stories. $\endgroup$ Apr 18, 2017 at 17:35
  • 2
    $\begingroup$ The problem with straight to the point is that you do not want them to seem simple and Tarzany. You want them to be complex but unsophisticated. +for the Darmok episode - I think of that too and that flavored this answer. $\endgroup$
    – Willk
    Apr 18, 2017 at 17:58
  • $\begingroup$ Straight to the point can include worlds of discussion that have no relevance to the nonnative speaker, therefore seem tangential and annoying. Crom cares not for these things! $\endgroup$ Apr 19, 2017 at 19:10

If you haven't read the book yet, speed read 'Sapiens' by Harari. He poses the idea that language developed amongst earlier humans not just to say "there's a lion at the stream, don't go there" but also to say "Jane's sleeping with Tarzan, don't go there either". In other words, gossip was as much of a driver when communities formed as warnings of external threats were (your displacement comment).

Think about the critical factors in the lives of the indigenous race and you'll find the function of their language. It might also be a very interesting way to build context in to their world by offering concepts which juxtapose to the readers' inbuilt prejudices or assumptions.


Now, I'm not an expert on proto-languages. But my immediate instinct is that this language should be as short and simple as possible. There are a few reasons for this:

  • As I understand it, your aliens are still at the "hunter-gatherer" phase. They might not even have invented writing yet - they may still be communicating via cave paintings. If that's the case, they'll be learning this language almost entirely through memorization, so the simpler it is, the easier it will be for them to learn.
  • There will be lots of things they haven't discovered, or concepts they haven't invented, or things they might not have come up with words for yet. For example, they probably won't understand the concept of "zero"; as I recall, that wasn't invented until a couple of thousand years after the first writing system. They might have a word for "small bird", but not for specific species like crow, pigeon, bluebird etc; they'd all just be "small birds".
  • If the language is geared towards hunting, you'll want to be able to communicate quickly and clearly, in as few syllables as possible. It might be the difference between catching your prey or going hungry. And if your prey is especially dangerous, it might be the difference between life and death.
  • It makes things easier for you. The less words and grammar rules you have to come up with, the easier it will be for you to create this language.
  • 1
    $\begingroup$ If they have numeracy at all, I like the idea that "hungry" and "zero" are the same word in their language. $\endgroup$
    – The Nate
    Apr 19, 2017 at 7:39
  • 1
    $\begingroup$ Its very easy to get the difference between lakes and oceans - you can drink the water from the lake, but not from the ocean. $\endgroup$ Apr 19, 2017 at 14:53
  • $\begingroup$ @JulianEgner That's a very good point. I'll change it to a different example. $\endgroup$
    – F1Krazy
    Apr 19, 2017 at 15:08
  • 1
    $\begingroup$ @JulianEgner Which doesn't necessarily imply that a language will have different words for them, however. See the history of the English word "sea" and its Germanic cognates, for example. $\endgroup$
    – DLosc
    Apr 20, 2017 at 8:53
  • $\begingroup$ @DLosc Thats true. In high German, its "Der See" for a Sea, but "Die See" or "Der Ozean" or "Das Meer" for the ocean. In the dialect spoken next to the north sea, a "Meer" is a Sea and "Die See" is the ocean. $\endgroup$ Apr 20, 2017 at 17:12

A lot of new languages start out using conjugations instead of coming up with new words to describe things. For example, instead of calling something a big fancy word like "bracelet", they'd come up with something simpler and more familiar like "arm-ring". My recommendation would be to keep speech very limited, maybe even just have the majority of their sentences be made up of one or two words ("get" and then point at something) and when they're forced to engage in more complicated conversation, use lots of conjugations, utilizing mainly contextual words.

As for phonetics, use lots of hard consonants and words with only one or two syllables.

  • 4
    $\begingroup$ Let's see... Thing Explainer and Uncleftish Beholding seem like things that should be on the reading list. $\endgroup$ Apr 18, 2017 at 18:22
  • 4
    $\begingroup$ Big word like bracelet... Actually, you see, what happened is that for historical reasons English uses a lot of words borrowed from an old and obsolete form of Frech. And in that form of French a bracelet was very regularly a small bracel; the word bracel being simply inherited from Latin brachiale (pronounced braciale in the mouths of ordinary people), meaning, very transparently, something to do with a brachium, an arm. Not such a big word at all, just an ordinary lexical borrowing from a language with which English was in intimate contact for centuries. $\endgroup$
    – AlexP
    Apr 18, 2017 at 20:59

Developing a completely new language (even the most simple one) is a tremendous task, it is very much easier to twist and extend an already existing one. Your goal language sounds very much like Toki Pona (https://en.wikipedia.org/wiki/Toki_Pona) for me. If you slightly changed its vocabulary and added a bunch of new words to describe concepts specific to your story, you would get exactly what you are looking for - a very simple but extendable language.


Toothed whales each appear to develop a signature whistle. So it appears that, at a minimum, they are constantly swimming around identifying themselves.

So it looks like in the simplest case, your language might just consist of self-identification, with any other information conveyed out of context.

To get a bit more complex, you could add sounds for basic verbs. You could perhaps hunt with this, by using Ogg's sound with the verb for "flank", to tell Ogg to go around while everyone else continues with a frontal attack.

To get more complex yet, add direct objects. So you can tell Ogg to use his club while flanking, while everyone else sticks to their spears.

Note that the more words you add, the more different sounds your primitives need to be capable of making (and discerning). That will require more flexible communication systems, and smarter creatures.


You should look into Noam Chomsky's theory on the origin of language, that it didn't emerge as a gradual iteration from other non-language animal communication systems, but that instead it arose very suddenly, from a freak change to the way human brains worked.

The idea is at first very counter-intuitive, especially if you are used to the idea of biological evolution, where features almost always develop incrementally, but there is significant evidence to support the theory. Principally, if it were the case that language is something that incrementally developed, we could expect to see that some present-day languages would be further ahead or further behind on the evolutionary ladder than others - but this is not the case. Instead we find that all documented languages seem to be equally advanced, yes, some languages have features that others don't, but the 19th century idea that languages could be placed on a hierarchy with "primitive" languages at the bottom, and with European languages (naturally...) at the top has been thoroughly discredited.

What is the case however is that languages do adapt to fit the cultures that speak them. An agricultural culture tends to have greater vocabulary about agriculture, while a hunter-gatherer culture would tend to have a greater vocabulary about hunting and gathering. I'd suggest that instead of trying to engineer something "primitive", you should think more about how to match the language to the conditions of the culture that speaks it.

(Or you might think that the Chomskyian idea is a load of rubbish, and that language did arise incrementally from non-language communication, in which case feel free to ignore this answer!)

  • $\begingroup$ Biological evolution does not lead to an "evolutionary ladder" with some species more advanced than others. This belief, like the one about languages, was based on 19th-century concepts of natural order and hierarchy. $\endgroup$ Apr 19, 2017 at 23:21

This is only one aspect, but it has not been adressed in any of the answers (or comments) yet:

Use simple present tense only.

No past, no future.

This is in line with what apaul3248 writes in his answer about learning a foreign language: you always start in the simple present tense.

It meets your requirements concerned with the immediate and no deep abstract concepts yet.


A thing to recall is that the language contructs available often largely dictate the thought patterns available to a society (and partially the other way around). Thus, if these aliens are truly only concerned with the immediate do not provide them with constructs that detail time. Any time into the future may be "tomorrow", while any time before now may be "in memory".

Further, it may be worth considering whether a codified grammar matters. In real life, few concerned themselves with grammar until laws came into place that had to be unambigiously interpreted. That may again not be a concern for these aliens.

Finally, if there is truly a large hunting aspect to these creatues; consider that their language may not be fully vocal. Sign languages were often used to not give away one's position, while whistling languages are very practical for communicating over much longer distances than shouting (comfortably) could.


I agree with keeping the language limited to one or two syllable words which are joined together to express more complex words. You should start by coming up with the most commonly used words like I/You/or/and/the and so on, building the language outward from there. Keep testing your words together in simple sentences to ascertain whether you like the sound of it, and if it captures the essence of the beings who created it.

  • 2
    $\begingroup$ That is, assuming that the language has articles (quite a rare thing in old languages) and personal pronouns on the model of English. What if pronouns distinguish genders (quite common)? And please please consider that prepositions are notoriously hard to translate -- it is exceedingly rare for two languages to have prepositions which correspond nicely one to one; for example, English "in" may correspond to French "dans", "chez", "en", "à", and of course each of those French prepositions correspond to several English ones depending on the specific meaning. $\endgroup$
    – AlexP
    Apr 18, 2017 at 21:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .