Pandorum is a generation ship movie. Generation ships are a fairly big sub-genre of sci-fi where people live for many generations in a ship while travelling to a distant star.

The movie Pandorum did not use a conlang at all, despite the ships residents devolving into a tribe of "primitive" natives. What sort of features would we expect to find after 1000 years of travel in space, using real world island languages as a guide?

  • 2
    I disagree. The only "feature" is that it would have diverged from the parent language significantly.
    – curiousdannii
    Commented Mar 2, 2018 at 1:13
  • 2
    Or perhaps that the parent language would have diverged from it, if you go the way of Icelandic's conservatism.
    – curiousdannii
    Commented Mar 2, 2018 at 1:46
  • 1
    Are we assuming that the original inhabitants of the ship only spoke one language, or several languages? I'd imagine that one lingua franca would be used initially, but I'm not sure.
    – HDE 226868
    Commented Mar 2, 2018 at 2:54
  • @HDE226868 The movie shows a multicultural ship crew. One of the actors was Vietnamese martial artist, but I don't recall him speaking. Everyone in the movie spoke ordinary English of course. Commented Mar 2, 2018 at 3:23
  • How many people in the ship? This would affect significantly what happens on board - not only linguistically. Commented Mar 4, 2018 at 23:31

14 Answers 14


The idea that language would not change over 1000 years of travel in space is absolutely ludicrous. 1000 years ago was before Middle English existed. Massive amounts of language change can occur over such a long period of time. Even very conservative languages change significantly over such a long period of time. No language will remain unchanged after 1000 years. Based on natlangs, it's more likely than not that the language will not be mutually intelligible with its predecessor after 1000 years.

How exactly the language changes depends a lot on the details of the situation -- how multicultural is the ship? What is the lingua franca of the ship? How are other languages viewed relative to the lingua franca? If there is enough of a divide in where different languages are used (such as if everything for official purposes is in English but a great number of native, say, Spanish speakers speak Spanish with their families at home), you may end up with diglossia arising. If there are significant numbers of multilingual people on these ships, you can bet there will be a lot of borrowing from minority languages into the dominant language, at least in less formal contexts. As in real life, there may be stigma against using such borrowings.

Honestly, language change in this situation won't likely differ that much from language change anywhere else. A conlanger has a lot of freedom in how they want to deal with the diachronics in this situation. The only real difference is that you don't have any chance of outsiders suddenly arriving and bringing with them waves of loanwords. ...Well, at least, you probably don't.

  • Comments are not for extended discussion; this conversation has been moved to chat.
    – curiousdannii
    Commented Apr 3, 2018 at 16:42

It's worth noting that widespread literacy, availability of written materials and public education are capable of greatly slowing down rates of change in the standard language. Additionally, if the generation ship's inside are designed with an ecology (as some SF ships are, like Bernard Werber's Le Papillon des étoiles or Léo's Centaurus), you can't reasonably expect a shrinking of vocabulary about nature. Basically, in a 1000 years I'd expect changes, of course, but I'm not sure we have the comparison levels to characterize just how much chances are likely to actually occur because the rates of change of standard languages over such long periods of time (presuming no "dark age" occurs somewhere along the line) are currently simply not really known.

Case in point: literary French and English from 300 years ago are still pretty comprehensible today (Voltaire's first play was presented that year, Dafoe' Robinson Crusoe is from 1719...). Go to three hundred years before that and you're firmly in Middle French/English, which are definitely rather different beasts.

  • 1
    True, but it's not terribly difficult for a modern, fluent English speaker to learn to read Chaucer. It takes some work, yes, but it's easier than learning a dead language from scratch. Commented Mar 2, 2018 at 15:50
  • 3
    Modern Icelandic speakers can apparently read the sagas from over 800 years ago.
    – curiousdannii
    Commented Mar 3, 2018 at 23:23
  • 3
    @RobertColumbia It's not terribly difficult to read, but that doesn't mean you could hold a conversation with Chaucer if you met him. A lot of spellings didn't change while the pronunciations of words and their use both syntactically and semantically did. Further, Chaucer wrote The Canterbury Tales less than 700 years ago -- 1000 years ago would be nearer to the time of Beowulf, which I guarantee a modern English speaker cannot read without learning the language from scratch.
    – Sparksbet
    Commented Mar 5, 2018 at 17:57
  • If, however, you had a recording of Chaucer that was used to routinely to instruct students and movies based on actors contemporary to Chaucer performing TV shows and movies based on Chaucer, and documentaries about life in Chaucer's time shot by Chaucer's contemporaries, the whole argument changes. Commented Oct 13, 2020 at 4:06

Vocabulary Changes

New words

The vocabulary can be expected to contain a few new terms or simpler ways of describing certain things that might be seen a lot or might be new, such as new star systems, stellar formations, etc. The words for some of these terms might already exist but shorter terms and compound nouns might come into play to make everyday speech in the Generation Ship a bit simpler.

Lost words

In addition to new terms being added to the vocabulary, some terms may go out of use and become lost from the vocabulary; for example, certain things that might only be seen on a planet (like a tree for example, or a mountain). Terms for those might be lost because they would likely not be used very often if at all. Once the ship arrives and these concepts are explored again, new terms might develop, though they might be similar if the terms are created based on modifying and combining existing terms (tree might be "tall bush" and mountain might be "giant mound", for example).

  • 6
    There is no justification for claiming that the phonetics or grammatical structure would not change over time in this situation because "there would be no reason" for them to change. That's not how language change works. Honestly, almost nothing in this answer reflects how language change functions in natural language.
    – Sparksbet
    Commented Mar 2, 2018 at 3:53
  • @Sparksbet Hm I suppose you're correct. Would my points about the changes to vocabulary make sense or should I delete? Commented Mar 2, 2018 at 13:18
  • Your vocabulary change ideas do make some sense, though I think it likely that many of those words would stick around for metaphorical meanings
    – Sparksbet
    Commented Mar 2, 2018 at 15:02
  • @Sparksbet Okay, thanks. I suppose that would make sense as the words may be used in situations even when the exact meaning of it is not represented ("strong enough to move a mountain", though that's more of a hyperbole, but I understand your point), but I believe my point holds that at least some words may go out of use or at least become very rare to see around. Commented Mar 2, 2018 at 15:27
  • Yeah, that's understandable! Your edit removes my problems with this answer.
    – Sparksbet
    Commented Mar 2, 2018 at 15:33

My guesses would be that it evolves towards one of two extremes:

  1. less morphology, fixed word order - this is what happened to English. Dropping cases and most inflections, but having a stricter word order to compensate for the loss of morphological markers.

  2. more morphology, freer word order - not sure if that actually is realistic. It's harder to invent new inflections than to drop them, and unless there is a flourishing poetry scene (perhaps to battle boredom), there wouldn't be too much reason for making word order less strict.

Most languages seem to navigate a path between those extremes, but that would mean it'd just become a different language altogether.

Looking at some real-life examples is tricky, as Latin -> French/Spanish/... for example had lots of interference from other languages and external events. In general I would think the language becomes simpler. It's a closed community, so changes can spread quickly and are not a hindrance to understanding. The environment is fairly static, so you don't need complicated phrases to refer to things.

One aspect would be communicating with the ship's computer (assuming a natural language interface) or texts. But you could think that a 'priest' class develops which does that communication and preserves the 'old' language. A bit like the Vatican with Latin.

Word meanings would probably change, as there are many words but only a limited number of things that need to be named; words would probably be re-used, or acquire new additional meanings.

Some things that won't change are the underlying distributional laws of language. There will still be a number of high-frequency short words that make up most of the active vocabulary. Zipf's law will still apply. The relationship between phoneme inventory and average word length will also remain: fewer phonemes will result in longer words. Short words will be more polysemous than long ones. But overall language will adapt to it's purpose, communication, in the given environment. If there are no trees, then the word tree might suddenly refer to an antenna. Or grass might be used to mean "an area of metal flooring that is slippery when wet".


More or less, the generation ship language will be a natural evolution of the languages brought in by the first generation (their common language probably being something similar to L2 English).

Some things will be spacy: They need a new system of orientation in 3D space with lack of gravity or artificial gravity at work, directions like up, down, east, north, west, and south are no longer meaningful. They will also develop some slang with respect to parts of their space ship and forget all the vocabulary about geographical features on earth.

EDIT: The space ship is probably operating with an air pressure that is significantly lower than on earth, corresponding to an altitude of 1000 or 2000 m. In this case, the space ship language may acquire ejective consonants (For a correlation between altitude and ejective consonants, see Caleb Everett's work "Evidence for Direct Geographic Influences on Linguistic Sounds: The Case of Ejectives" http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0065275 )


I'll assume here that the generation ship in question's mission is a resounding success: the inhabitants were not attacked by huge insectile aliens that enslaved them all; the Computer did not rise up and shut out its erstwhile masters, leaving them to fend as best they could in a decaying piece of technology. No, society not only survived but thrived and the people who set foot on their new homeworld have at their command all the accumulated knowledge and technology they brought with them.

In this scenario, there will, I think, actually be very little change in the language spoken on board the generation ship. Whatever interlanguage the first generation brings aboard (likely English) will be the language of general communication among all the inhabitants of the ship. Education will be carried out in English, though programmes will be in place to teach "ancestral" languages brought on board as well. And of course, the homes of firstgen folks will resound with their own native languages.

As subsequent generations are educated (in English), familiarity with other languages will fade. Since this project, while still Earthbound, brought together the best and brightest and well educated to begin with, their level of comfort with English will be high regardless of their cradle tongues. Because running a generation ship requires a huge commitment in education (sciences, mathematics) & experience (astrogation, complex technological subsystems), communications skills will be highly prized. Just as with Modern Standard American, the nature of life & culture aboard ship will tend to keep a fairly strict standard language in place. We're basically looking at a situation where everyone is highly literate and most likely cross-trained in several disciplines ánd certainly likely trained in one or more "traditional" crafts, artforms or skillsets.

This is not to say that jargon won't arise or that English will all of a sudden stop borrowing words from other languages! Far from it. By the time we get to journey's end, I'd speculate that lastgens will be quite able to understand the recorded speech and song of firstgens and that their languages will be not much different at all.

  • 3
    As a European scientist in Japan, I greatly disagree with the idea behind the third paragraph. Just because people are very smart and excel in at least one academic subject does not mean their understanding of English is good or they are confident using it.
    – Jan
    Commented Mar 2, 2018 at 11:25
  • 2
    No worries, and I understand where you're coming from! Thing is, on Earth, we can have a dozen countries with highly educated scientists etc that don't speak English well. In a project like this, the onboard community really doesn't have that luxury. Everyone needs to be able to communicate at a high level of competency. Regardless of the languages people bring aboard, there will likely be some kind of convention interlanguage. I chose English as the most likely. It could be French or German or Japanese. Whatever it is, that's the language everyone's going to be speaking by journey's end!
    – elemtilas
    Commented Mar 2, 2018 at 23:31

A generation ship is a small society that is technologically advanced but stagnant. It's a society that depends on ancient wisdom to survive. Their material and intellectual resources are very limited compared to the civilization they came from, so the rate of scientific and technological advancement will be slow. Most of their efforts in those fields must go into maintaining a sufficient understanding of the ship and the science behind it to keep the systems operational.

For similar reasons, most of the entertainment the people consume will be from the civilization of their origin. Some new entertainment will be produced, but in limited quantity and with lower production values.

Because of limited resources, I would expect that everyone on board would be bilingual. There would be an everyday language that would drift and evolve, as others have already discussed. There would also be the ancient language of entertainment, higher education, and operations manuals that would remain static by necessity.

Bilingual people tend to mix the languages when talking with each other. This would make the evolution of the everyday language different from what we have seen on Earth. Because the original language would still be in active use, the everyday language would sometimes drift away from it and sometimes be influenced by it.


Others have mentioned that vocabulary for things not seen in space might vanish, and @HyperNeutrino mentioned that new terms might arise if the ship reaches a planet.

New terms might come into existence as metaphorical references to things that the colonists were familiar with on the ship. Just like we refer to "folders" and "files" on a computer or speak about "surfing" the "net" (using older terms to metaphorically refer to newer things that needed terms), generation ship colonists newly landed on a planet might refer to a land vehicle as a "shuttlecraft" or "capsule", call their political leader a "captain", or refer to the death penalty as "being thrown out of an airlock" (even if the actual means of execution becomes something else). Today we sometimes refer to uploading files as "posting" them. Our colonists might rediscover physical message boards but refer to posting notes on them as "uploading".

There could be a wider cultural effect too that expands to include both linguistic and non-linguistic features. If colonists are used to a shipboard timekeeping system that is detached from astronomical phenomena, or that is based on astronomical phenomena only visible from elsewhere in the universe, they might continue using it after landing on a planet despite the presence of astronomical phenomena that could be used to construct a calendar. Consider how the Apollo astronauts on the Moon continued to use Earth-based timekeeping metrics even though they could have adopted some sort of "Moon time" based on the cycle of the Lunar day and/or the apparent rotation of Earth in the sky. This sort of phenomenon could cause holidays to "rotate" throughout the seasons of the new homeworld in a similar way that dates defined in Earth-based lunar calendars (e.g. Ramadan) can occur in any solar season. Perhaps the new colony world has a year that is half that of Earth, so Christmas now comes every other northern hemisphere winter, or the year is twice as long as Earth's, so the colonists celebrate Christmas at the beginning of winter and again at the beginning of summer.

See Robert Heinlein's Orphans of the Sky for some examples of how this might look.


The particular circumstances on a generation ship will work in several directions. Some accelerate language change, some slow it down.

  • Accelerating: Language is used to create identity. The first, most important and biggest tribe are the ship's crew and passengers (if such a distinction is made). They will start to set themselves apart through names for themselves versus the planet dwellers.
    Within the ship, different sub-tribes like engineers, gardeners, educators etc. will create respectively continue to develop their own slang in order to set themselves apart from other groups. The urge to set themselves apart is greater because the ship packs many people in a comparatively small space, creating a need to draw virtual borders.
  • Accelerating: Groups with lots of spare time (which I suppose our space travelers will have) generally devote some of it to developing sophisticated tastes, fads and fashions, thus creating in-groups; their language evolution is interacting with these identity functions. Individuals who are unable or unwilling to speak "their slang" (and wear their hair cut, and posture ...) are marked as "not belonging to them". To a degree this happens in every generation of teenagers. But we can see that also on a larger scale in the development of the aristocratic French, or the fast-paced development of the politically correct speech in the United States.
  • Slowing down: There will be constant communication between all individuals of the ship through some kind of publications and messaging as well as personal conversation, given the necessary physical proximity and the need to cooperate during the mission (no group will be able to isolate themselves). This will ensure that there will continue to be a single lingua franca, fads and fashions not withstanding. Some people may learn and teach other languages for scientific, identity, hobby and nostalgic reasons, but they all will be competent in the main language. An important driver of the evolution of this language will be the slang of the youngsters, as always.
  • Slowing down: The generation ship will probably communicate via radio with the planet they came from; but most communication will be internal, partly because after a while communication with the home planet will not be real-time any more. The generation ship will form an isolated, small community. Such communities, like the Amish in America or the German settlers in Romania, are known to preserve customs and language longer than the bigger communities thy split off of. We can assume that the ship receives entertainment from the home planet, like music, movies and written prose. Thus they will keep in touch with the language on the home planet; whether that ongoing contact is sufficient to keep the language on board from running its own course I don't know.

I recently read an interesting essay – wish I could remember where – saying that the more insular a language community is, i.e. the less likely someone is to converse in that language with a stranger, the less it uses syntactic complexities such as relative clauses, because you don't need to clarify your context if your interlocutor already shares most of your context.

If that's true, the effect is likely to show if the ship's population stays within a small multiple of the Dunbar limit. Such simplification might in turn allow word order to become more free, I think.


It depends. There are two extreme possibilities. I would expect the real situation to be somewhere in between.

One extreme happens when the population of the generation ship just speaks a random language, or several. And when it either does not bring a large library or somehow manages to lose or forget it (e.g. due to religious extremism).

In this case there will be natural language change. Here are some things that fuel language change:

  • The desire of a younger generation to assert their identity linguistically gives rise to youth language. As they grow older, part of their youth language becomes standard.
  • Unstressed syllables are pronounced less and less carefully, giving rise to phonetic changes. E.g., several distinctive vowels may merge into a single neutral vowel when unstressed. Vowels or consonants may even get lost.
  • The desire to express ourselves drastically in certain situation gives rise to new rules of grammar. E.g., this is how the future tense and the progressive aspect in English arose. (Neither existed in Proto-Germanic.)
  • As a word or expression is used increasingly in a specific sense, sometimes it is no longer appropriate to use it in its original sense. This drift of meaning then creates a gap that will likely be filled by a new word or expression or by initiating a similar process for another, existing one.

All of these factors exist on a generation ship. As ohwilleke pointed out, two other factors - changing environment and contact with other languages (which also evolve) do not exist on a generation ship. (At least in the long run. Of course the first few generations will adapt their vocabulary and probably even grammar to the unique situation. E.g., some Australian languages use compass directions or geographical directions w.r.t. a specific valley instead of left and right. Something equivalent would make even more sense on a generation ship.) Therefore we can expect this language change to be really slow. Ohwilleke's example of Iceland gives us an idea of what to expect. If the ship's population starts with English, then after 1000 years they have a good chance to understand today's English to some extent. (To a greater extent if they continue reading books in today's English.)

Since, realistically, a generation ship to Alpha Centauri would be travelling for many thousands of years, without further stabilising factors we can expect that they would arrive speaking a language that is at most vaguely similar to the one they started with. I would not expect this language to have unusual features compared to languages currently spoken on Earth. There might, however, be some grammatical structures related to the special situation on the ship. And there might also be a large number of irregularities, as is often the case for languages of small speaker communities.

A generation ship will probably bring a large library. However, if the population had unrestricted access to it, this might lead to a society that basically lives in the past and will be unable to settle a new planet. But in any case I would expect that some people on board continue to learn Earth languages so they can understand part of that library. Depending on the precise circumstances, this could be a stabilising factor or even fuel language change due to different Earth languages being fashionable at different times.

A civilisation that wants to control the language spoken by the settlers of a distant planet will have to turn linguistics from a descriptive science into a field of engineering. They may be able to create a language that is a lot more stable than any naturally occurring language. This doesn't mean there is no change at all, but it means that significant change is normally cyclic.

Language change is already largely cyclic in terms of high-level features of languages. E.g., if a language does not have a past tense, then a composite tense like present perfect will likely come up to fill this gap. The auxiliaries will then merge with the verbs that carry the meaning, becoming suffixes or prefixes. They will slowly lose stress, become unrecognisable as auxiliaries, become indistinguishable from each other and finally will disappear, so that the language again doesn't have a past tense.

For an isolated speaker community, it might be possible to come up with a language A such that after one such cycle a language B results that is mutually intelligible with A to a strong degree. Now the speakers of B will likely read a lot of the literary output of the speakers of A, leading to further stabilisation.

Extending this idea, we one might come up with a family of languages that could in principle have cyclically evolved out of each other. Say half a dozen such languages. By providing an extensive and stimulating library in all of these languages, it could be possible to make language evolution essentially cyclic throughout the flight, so that at the end everyone will be speaking a variation of one of the library languages (possibly an intermediate stage). But it also seems likely that this would lead to a multilingual situation (with several library languages spoken simultaneously), so that one would also have to consider contact effects between these languages. And, perhaps more importantly, to conscious efforts to create linguistic innovations that are not represented in the library.

Note: I arrived at these last thoughts from a different angle. In the extreme case, success of an international language such as Esperanto would mean that everybody in the world speaks Esperanto and only Esperanto. But this is not a stable situation, at least not with an essentially random language such as Esperanto. So I thought about whether there can be a language that would remain the only language spoken on Earth once it managed to get to that point.


I found a (rather new) publication on this theme, namely

McKenzie, A., & Punske, J. (2020). Language Development During Interstellar Travel. Acta Futura, (12), 123–132. DOI 10.5281/zenodo.3747353

The publication is rather high-level invoking analogies to the Polynesian explorers, the Malagasy language, and the Balkan Sprachbund.

I became aware of this by the following blog post:



I'm sort of bemused that every answer I've seen has seemed to overlook the rather large elephant in the room.

Data is among the cheapest, most storage efficient things you can carry. Plugged into my computer at this moment is a 512 GB thumb drive. It is, in its (thick) protective case, 50mm long, 20mm wide, and 10mm thick.

This is a generation ship, made by a technologically advanced culture. Let's assume, for simplicity and practicality's sake, that everyone is fluent in one common language. Just for the hell of it, let's say English. And so, to keep people entertained, not only written material is provided but recordings. Lots and lots of recordings. Music. Movies. Documentaries. Learning material. TV shows.

Assume they provide the material in 720p resolution, so not quite full HD but better than standard. You'll get about 3.5 hours per gigabyte. Rounding off at 500GB of data, my perfectly ordinary thumb drive has a capacity of roughly 1,750 hours of audiovisual recordings. That's over 10 weeks of video. That's every Marvel Cinematic Movie released so far repeated 35 times. That's 2,500 "one hour" (ie, 42 minutes) North American television episodes. On a single drive.

There is a box beside my desk that is abut 100x150x300mm in size. I could stick about 450 of my thumb drives in there. That's 225,000GB, or about 787,500 hours of video. That's ninety years of video. All in something the size of a small box easily picked up by one hand.

So, one thing we could safely assume, if no one is trying to censor the material for some reason, and putting aside the stability of storage media, since if we're talking a long duration ship and everything is still working is probably a mostly solved problem, the people on board may well have access to a ridiculous amount of recording of people talking, and they're going to grow up with it.

There's not going to be any argument over how people talked a 1000 years before. No equivalent discussions over whether someone pronounced vici as "veechee", "weechee", "veekee" or "weekee" because they will know. They'll know about different accents and how people pronounced vowels and consonants and how the written was different from the spoken because they can literally see and hear it first hand. It is what they will learn from because they grew up with it, just as their parents learned it, and their parents, and their parents.

Everyone proposing the language will radically change in this scenario, how?

You can't use a historical example of how much a given language has changed over a few centuries to make your argument because you are ignoring the fact that audiovisual recordings fundamentally alter how people communicate and never existed during the time spans your examples are going to cover. You can't cite historical precedent regarding language change because there isn't any. The situation we've been in over the last century and a bit in terms of communication is entirely unprecedented in the history of the freaking planet, let alone our species.

  • I don't think that the preservation of audio records will keep language from changing. The spoken vernacular on the space ship will soon develop an accent different from the "classical pronunciation" of the historical records. One can compare this to frozen pronunciations in religious languages with detailed prescriptions.
    – Sir Cornflakes
    Commented Oct 13, 2020 at 10:00
  • 1
    I grew up watching movies made between the 1930s and 1950s. I always wondered about their funny accents. Modern Japan in this century has mass media & the older generation likes to complain about how much language has changed. Commented Oct 13, 2020 at 12:44
  • I'm not saying accents won't change (if due to fashion, if nothing else), and there may well be changes in slang, but people are speaking as if the language will be incomprehensible after 1000 years, and no one can explain why that would be. Commented Oct 13, 2020 at 15:01
  • Within a millennium or two, even this will happen. Classical Arabic and Sanskrit are both incomprehensible to the speakers of their descendent languages without special education.
    – Sir Cornflakes
    Commented Oct 15, 2020 at 16:17
  • 1
    Both for Classical Arabic and Sanskrit there are detailed prescriptions on how to pronounce each letter in those languages, and trained reciters of the religious texts are present delivering that classical pronunciations to the public. Nonetheless, Arabic vernaculars and modern Indoaryan languages have evolved to be different from that.
    – Sir Cornflakes
    Commented Nov 16, 2020 at 11:41

Short Answer

The language used on a generation ship would change very little over the course of the journey despite its 1,000 year duration.

Long Answer

Contrary to widespread conventional wisdom, language drift is very modest over long periods of time in the absence of specific environmental or language contact experiences the cause a language to change over time.

One of the best historical comparisons is the Icelandic language. It is almost unique in experiencing very little language contact or mass migration for a very long period of time. (In contrast, for example, Old English arose from the Anglo-Saxon invasion of a linguistically Celtic and Latin speaking island, and Middle English arose as the main modification in the wake of England's conquest by a French speaking Normal elite and England was extremely involved in contacting other civilizations globally as part of its diplomatic, trade and cultural interactions with Europe and as a result of its global empire.)

Iceland was an uninhabited island (hence not subject to any substrate influences unlike many other colonists) settled in the historic era by colonists who spoke a single shared language fairly close to "Old Norse" (the proto-language of the Germanic languages), who had a total population similar to that of a generation ship.

Somewhat more subtly, Iceland's colonists were looking for economic opportunities, rather than constituting a cultural minority fleeing persecution in their homeland the way, for example, that the Puritan colonists of New England were, so they were not motivated at the outset to intentionally set out to put linguistic and cultural distance between themselves and the speakers of their original mother tongue the way that early Americans deliberately attempted to distinguish themselves linguistically from England to assert a distinct cultural identity. The Iceland situation of a group of people not committed ideologically to putting cultural distance between themselves and their ancestor population would likely also be true of the people on a generation ship.

The descendants of the original colonists of Iceland remained in constant communication with each other ever after, just as generation ship residents would, although this did break up from a single community into about a dozen or so chiefdoms that were substantially autonomous from each other and had distinct identities although they did retain a very thin form of island-wide weak government. So, tribal divisions, per se, do not distinguish a generation ship from Iceland.

And, Iceland received almost no significant waves of migration prior to the recent past, so it has never had a large community of outside language learners to influence its linguistic development (just as would be the case in a generation ship).

While Iceland wasn't as isolated from other people as generation ship would be, for a period of time roughly equal to the length of a generation ship journey prior to the introduction of telecommunications, Iceland was very isolated from the rest of the world, and the vast majority of what little contact it did have with the rest of the world from visiting ships was with people who spoke the closest linguistic relative of their language in the world. So, it did not have any languages from which to borrow words, and it was not subject to any areal effects from neighboring languages in a significant way (apart from Danish rule whose effect is noted below).

Icelandic is by far the most linguistically static of the Germanic languages, with the only significant change being a change in pronunciation that arose during a several century period of Danish rule, a country that speaks a closely related by different language derived from Old Norse. A generation ship would not experience this outside influence and would not need to develop a new written language as Iceland did.

As Wikipedia explains (at the link above):

The oldest preserved texts in Icelandic were written around 1100 AD. Much of the texts are based on poetry and laws traditionally preserved orally. The most famous of the texts, which were written in Iceland from the 12th century onward, are the Icelandic Sagas. They comprise the historical works and the eddaic poems.

The language of the sagas is Old Icelandic, a western dialect of Old Norse. The Dano-Norwegian, then later Danish rule of Iceland from 1536 to 1918 had little effect on the evolution of Icelandic (in contrary to the Norwegian language), which remained in daily use among the general population. Though more archaic than the other living Germanic languages, Icelandic changed markedly in pronunciation from the 12th to the 16th century, especially in vowels (in particular, á, æ, au, and y/ý).

The modern Icelandic alphabet has developed from a standard established in the 19th century, primarily by the Danish linguist Rasmus Rask. It is based strongly on an orthography laid out in the early 12th century by a mysterious document referred to as The First Grammatical Treatise by an anonymous author, who has later been referred to as the First Grammarian. The later Rasmus Rask standard was a re-creation of the old treatise, with some changes to fit concurrent Germanic conventions, such as the exclusive use of k rather than c. Various archaic features, as the letter ð, had not been used much in later centuries. Rask's standard constituted a major change in practice. Later 20th-century changes include the use of é instead of je and the removal of z from the Icelandic alphabet in 1973.

Apart from the addition of new vocabulary, written Icelandic has not changed substantially since the 11th century, when the first texts were written on vellum. Modern speakers can understand the original sagas and Eddas which were written about eight hundred years ago. The sagas are usually read with updated modern spelling and footnotes but otherwise intact (as with modern English readers of Shakespeare). With some effort, many Icelanders can also understand the original manuscripts.

Also, usually new vocabulary is developed in response to new topics of conversation. Since the environment would change less and there would be fewer new things on a generation ship than in Iceland over the same time period (in part, due to a generation ship's lack of trade with the outside world that Iceland had) one would expect there to be less innovation in vocabulary on a generation ship than there was in Iceland. Still, as noted by @Hyperneutrino one might expect words that residents of a generation ship have no reason to use to be lost early on from popular speech (although presumably written sources would still preserve the old words for those so inclined, as a generation ship's residents would not be illiterate and written languages stabilize spelling and prevent old words from dying completely), and there would be a few new words developed over time.

Another clue with which I have first hand experience, which corroborates the example of Iceland, is the development of motherland languages in immigrant communities. For example, the dialect of Korean spoken in Korean immigrant communities in the U.S. that formed shortly after the Korean War is much more similar to the dialect spoken at the time of migration than the dialect of Korean currently spoken in South Korea, notwithstanding the fact that there have been many visits to South Korea and there has been access to modern Korean television and music in that time period in the U.S. Small, relatively isolated communities are simply much less linguistically innovative than large communities that are strongly connected to neighboring communities that speak different languages.

In summary then, it is likely that the language of a generation ship would be extremely static over a period of approximately 1,000 years, because generally speaking it would be in conditions very similar to those of Iceland which historically has had an extremely static language over a similar time period under similar conditions. And, several of the factors that are know to have led to the modest changes between Old Icelandic and Modern Iceland over that time period would be absent on a generation ship.

  • Icelandic has been far from static though. Its phonology has changed greatly since Old Icelandic. Most of its vowels have shifted; long vowels broke into dipthongs and vowel length is now determined by stress rather than phonemically. Consonants did funky stuff too, and basically none of it is encoded in the writing system. The conservativeness of Icelandic is often very overstated; and on top of that it’s also undergone a culturally driven “reversal” to “proper icelandic” words, trying to replace loanwords with native coinings. One can say it’s been artificially turned older by this process. Commented Mar 3, 2018 at 21:52
  • @Adarain I think that there is a good case to make that the phonology changes are due to Danish influence during 300ish years of Danish rule. And, I suspect that most communities undergo waves of culturally conservative shifts from time to time. A generation ship's people might do that as well as some point in time.
    – ohwilleke
    Commented Mar 4, 2018 at 17:34
  • @ohwilleke - I think there is a better case to make that phonological changes are rarely due to external "influences". On the contrary, they seem to be overwhelmingly internal. Just look how loan words are always pronunciated according to the phonology of the borrowing language. Commented Mar 5, 2018 at 0:14
  • @Luis Henrique By their very existence loan words are an external influence. And changing the pronunciation of them has a practical value, especially if it contains sounds that aren't present in the host. Like a lot of English people I can't do the Welsh ll, for example. That's entirely different to me waking up tomorrow and pronouncing "chair" as "shayeer". Commented Mar 5, 2018 at 8:52
  • @LuísHenrique Many phonological features, such as the use of tone in a language, are areal effects that cross-language boundaries and are not obviously substrate effects.
    – ohwilleke
    Commented Mar 5, 2018 at 20:25

Not the answer you're looking for? Browse other questions tagged or ask your own question.