When writing past forms of English, you can (at least in theory) get everything right. There is no aspect of written language, from rhetorical style and syntax to the usage and even the spelling of single words, that you cannot check for accuracy against abundant primary sources and secondary literature.

But when it comes to the language of the future, there is no such thing as being right: there is only being plausible, which is much more subjective. That subjectivity doesn’t let you off the hook; readers will still call you on your slips. But it makes avoiding those slips more difficult, and it makes research less a matter of looking up specific details and more a matter of careful design.

Alas, there are no hard and fast rules for you to follow. Like economists, linguists are much better at explaining the past than at predicting the future; no one can say how English, or any other language, will change over the next fifty years, or the next five hundred. But its changes, whatever they may be, will be consistent with the trends and constraints that have shaped linguistic change in the past.

Here, then, is an overview of some of the most salient of those constraints and those trends. I don’t cover conlanging (language construction) in this article - if your story is set so far in the future that the characters’ language, even when written, would be incomprehensible to modern English speakers, you are back into conceit-of-translation territory (which was covered by the previous article), just as if you were writing characters who spoke any foreign language. However, forces that come into play when designing a new language also pertain to modifying an existing one, and so you may find conlanging resources to be helpful to you: to start out, try The Language Construction Kit, the CONLANG list, Pablo Flores’s Language Creation page, and the SIL glossary of linguistic terms, which may be useful if your research takes you into the linguistics literature.

WORDS: Lexical and Semantic Change

When we think of linguistic change, it’s lexical change that tends to come to mind first - new words are coined; existing words fall out of use, or come to mean something broader, or narrower, or altogether different. Lexical change is easy to show, and easy to make understood; it’s one of the most useful - indeed, indispensable - tools in the SF writer’s toolbox. But it’s also one of the easiest to misuse, or overuse.

In the vast majority of cases, a single question will tell you whether you’re justified in inventing a new word: Am I inventing a new thing? If the answer is yes - and if you are writing SF, chances are it often will be - then you’re on safe ground inventing a new name for it. Not, of course, that every new concept and gizmo needs a new name; if you want to quickly immerse readers in the story without calling attention to less important parts of the milieu, you have every reason to call the personal grav-resist thrust pods "cars" and the sub-vocalizing otolaryngeal implants "phones."

And not every new thing needs to be a new thing: James Blish railed against the tendency of SF writers to "call a rabbit a smeerp," to rename things for fantastic settings without changing their basic attributes. Color, provenance, and the presence of antennae are not basic attributes; if your characters drink a hot, bitter, caffeinated liquid brewed from dried leaves, they’ll probably call it tea even if the leaves are blue and come from the third moon of Zebulon Prime.

If you can do a search-and-replace for your coinage and make a mechanical, one-for-one substitution of a well-known English word without making any other changes, then your coinage is probably a smeerp, and the well-known English word probably the best choice.

But there is an exception to the smeerp rule, one circumstance in which people do simply substitute new words for existing ones, with little or no attendant change in meaning: when they do not want to be understood by listeners who aren’t in the know. Speech filled with this sort of one-for-one substitution is an argot. The most famous argot in spec fic - and one of the greatest pitfalls awaiting the linguistically-unwary sf writer - is the Russian-derived Nadsat of Anthony Burgess’s A Clockwork Orange.

Burgess’s novel may be taken as a textbook in how to introduce unfamiliar language quickly, thoroughly, and without interrupting the story. It should not be taken as a manual in how to create a near-future language unless your characters have as good a reason as Burgess’s Alex for speaking in code. Remember that in Orange, only the teens - more than that, only the teens from Alex’s own subculture - speak Nadsat, or indeed understand much of it. It’s an underworld cant, not general parlance.

Argot may and often does overlap with slang, and lends terms to it - witness Cockney rhyming-slang, which has shifted almost completely from the first category to the second. But while slang may be as impenetrable as argot, its purpose is subtly different - not to confuse outsiders, but to mark insiders. If argot is a secret code, then slang is a verbal secret handshake - it shows that the speakers are part of the same social group. But both slang and argot are subject to rapid change, as useful words pass into general parlance, and so lose their cachet or their secrecy.

The terms most likely to make that leap, however, will not be the simple substitutions for well-known terms - they will be words that express some meaning not easily stated by the standard language. So, for example, of the 17 slang nouns the OED lists as first cited in 1930, only five are still in anything approaching common use: moxie, mushmouth, Mickey Mouse (in a descriptive sense), gangbuster (though the phrase "come on like gangbusters" would not appear until the premier of the radio drama Gang Busters six years later), and dong. Only the last of those has any precise equivalent - and it appears to be a universal rule that no language can have too many words for "penis."

In short, though breaking the smeerp rule will sometimes be justified, it will never be justified for all characters in all contexts: they will not all use slang or argot; they will not all use the same slang or argot, unless they are all part of the same subculture; and those who are able and entitled to use the speech of a particular group may not use it with outsiders.

GRAMMAR: Morphological and Syntactic Change Most of the words in English, or in any other language, are content words - nouns, verbs, adjectives, and some adverbs. The content words are an open category: new nouns, verbs, and descriptors can be easily coined or borrowed and added to the language.

Not so with function words: pronouns, prepositions, auxiliary verbs, conjunctions, articles, and all the nearly-meaningless grammatical scaffolding that holds a sentence together. Change, in these categories, is very slow, and innovation very rare - new function words nearly always arise from some existing content word or phrase, as with the modern English auxiliary verbs evolving from the OE verbs I discussed in Part One of this article, or with the prepositions "atop" and "beside" which arose from the complete prepositional phrases "on top of" and "by the side of."

If you want to coin new English function words, there are two ways to proceed. One is to listen very carefully to the way people speak, particularly the words they tend to contract together and to never stress; to read up on syntax; and to create your new function words from semantically bleached words and phrases of modern English.

The second option, which I strongly recommend, is to not bother.

Note that this applies to pronouns, which are a classic example of a closed category. New pronouns may arise from existing pronouns, nouns, or noun phrases: so, for example, English y’all and yinz, from "you all" and "you ones," and Spanish Usted from vuestra Merced, "your Mercy" (used as a term of address like "your Highness"). New pronouns cannot be coined like new content words and introduced by fiat; of the many gender-neutral pronouns proposed in the last century, only zie/hir has achieved any currency at all, and then nearly entirely online; no one uses it in speech who has not learned it in a text-based forum, or speaks it without a degree of irony.

If it were much more often necessary to speak of sexless animate beings, or of a third biological sex, it’s likely enough that English would come up with an appropriate pronoun; and it’s likely, too, that such a pronoun might then spread to become the default. In the absence of a much larger category of non-male, non-female beings, however, it’s very unlikely that a new gender-neutral pronoun will gain any ground in English. English already has a singular gender-neutral pronoun, widely used in literature from Chaucer to Shakespeare to Jane Austen: they. Singular they/their has been around since Early Modern English, and even the pedants who rail most strongly against it still slip up and use it from time to time. If anything at all is certain about the future of English, it’s that singular they isn’t going anywhere. Embrace it.

In addition to content and function words, there is a third, more nebulous category - phatic expressions, interjections, particles, profanity, and frozen phrases of social pleasantry. Though many such words are largely meaningless, scaffolding for the conversation as the function words are for the sentence, they are an open category. Wide open, in fact; these words are among the likeliest to be borrowed from other languages.

Borrowing, of course, is a very common source of new words, and liberally peppering conversations with loanwords is a quick and plausible way to show cultural contact and influence - Maureen McHugh used Chinese to excellent effect in China Mountain Zhang, as did Joss Whedon, in a very different way, in Firefly. Do not assume, though, that after the greetings and stock phrases - the phatic expressions - the next words adopted will be the simple nouns and verbs next to them in the first chapter of the language primer. They won’t be. Instead, look for anything that can stand on its own as the answer to a question (phrases like "no problem," "enough," "whatever"), for idioms that don’t translate well, and for anything with no exact analog in English. Historically, terms for attitudes and states of mind have been much-borrowed (Weltschmertz, sang-froid), as have terms of art (wabi-sabi, the entire Italian musical vocabulary) and words dealing with food.

Besides borrowing and coinage ex nihilo, new words may also be created from existing words - through compounding, by the addition of prefixes and suffixes, or by the appropriation of words from one part of speech to another. This sort of word-formation is the most common, and also the most invisible - if the meaning of the word is transparently derivable from its parts and its structure, readers and hearers are unlikely to even notice that the word itself is new to them, unless the word is exceptional in some way - in length, aptness, or whimsy.

Such morphological processes are not only a source of neologisms; they can also be turned to good stylistic effect. Consider the highly distinctive banter of the characters on Buffy: The Vampire Slayer. Buffy and her friends never borrow or coin words in the same way as Alex and his droogs; their innovations are almost entirely morphological and syntactic. They create compounds ("Maybe you should be fray-adjacent"), turn adjectives to verbs ("Think you can vague that up for me?"), noun phrases into adjectives ("How very serial killer of you"), and names into common nouns ("You’re the useless part of the group! You’re the zeppo!"). They may unpack words into new, complex constructions ("Did anyone else just go to a bad visual place?"), or let single words stand in for a whole clause ("Raise your hand if ewww.")

None of these processes are unique to the Buffy characters’ way of speaking. The rules of word-formation and usage they exploit are common to all English speakers, though they are seldom applied so broadly and so frequently. The cumulative effect is one of great innovation, but with no new rules or processes - and so requires little effort to understand.

Attempts to show grammatical change above the level of individual words, at the large-scale structure of clauses and sentences, tend to quickly reach a point of diminishing returns. Haphazard change is jarring - if like Yoda your characters talk, laugh at you the readers will - but designing systematic change is a major conlanging project. And that project is really only worth undertaking for the sheer geeky pleasure of it - if the resulting grammar is close enough to that of modern English to be easily understandable, the payoff in distancing and exoticising won’t be worth the effort.

Rather than systematically overhauling English syntax, you would do better to exploit the syntactic variation of modern English, and its wide range of quirky, marginal, or frowned-upon - but completely understandable - grammatical devices. Keep a notebook of the phrases your friends only utter when drunk or on the phone with family - "Oh, for cute" or "I do it all the time anymore" or "So don’t I" - and play with expanding the constructions: So haven’t I; So didn’t she; so won’t we.

A final word on the subject of syntax: there are a number of grammatical changes currently in progress in English which are likely to continue - for example, the loss of case distinction in compound phrases, or anywhere outside the subject position of the sentence, which has given rise to constructions like "It’s me" and "Sarah and me are going out." It’s quite likely that in another few generations, such constructions will be completely unmarked in English, and proscriptive rules against them will seem as quaint and unmotivated as the rule against ending a sentence in a preposition.

You are writing, however, for a modern audience, and such up-and-coming constructions can still arouse a great deal of prescriptivist ire. If such reactions works against your characterization - if you mean your characters to come off as highly erudite - you may be best going with what modern prescriptivist grammar dictates to be the correct form, however unlikely it may be to endure.

SOUNDS: Phonological Change There is a persistent myth that language evolves toward greater simplicity. But there is no single definition of what could constitute simplicity in anything so large and necessarily complex as a language - and if there were some optimal, perfectly simple linguistic ideal, surely some language would have found it by now.

Neither, however, is language change a matter of degeneration and decay. Rather, language evolution is best thought of as a continuing conflict between the needs of speakers and hearers. The hearer wants redundancy, every piece of information encoded several ways, to make sure nothing is lost; the speaker wants to say things once and once only. The hearer wants adjacent sounds to contrast greatly, so as to be easier to distinguish; the speaker is concerned with ease of articulation, and wants adjacent sounds to be similar.

For any language, at any time, some of these conflicts will be settled in the speaker’s favor, and some in the hearer’s. But since each language user plays both roles, such settlements are never stable, and there is no optimum. Change is a constant. Phonological change - change in the sound system of a language, where this conflict plays out on the very small scale - is not only constant, but often quite rapid. It is certain that in any language group, new accents will arise.

What is far less certain that you will have any need to call attention to them. If all your characters speak with the same accent, there is no need to describe its particulars. If an accent is noticeable to your point of view (POV) character, then by all means remark upon it - but be as wary of attempting to show it orthographically as you would be of writing dialect in modern English. If the accent is also characterized by distinctive word choices or grammatical quirks, those details are quite enough to fix the sound in the reader’s mind; a sentence or two of description is all you need to tell us about the vowels and consonants, the speed and the stress pattern. Attempts at phonetic spelling are almost never justified.

Even if a character’s accent is nearly or completely impenetrable to your POV character, it’s almost certainly overkill to write out what your narrator hears. Stick with what she understands, and what she thinks of the accent; it’s what’s in her mind that your readers care about, not what’s hitting her eardrums.

As with syntax, major overhauls to phonology should only be undertaken if you’re into conlanging for its own sake; however you decide your future English is pronounced, you’ll still need to spell it in a way your readers can interpret. (Iain M. Banks and Russell Hoban are professionals on a closed course: do not try this at home.)

THE LINGUISTIC ENVIRONMENT: Social and historical factors. Language does not change at a constant rate. Three hundred years separate us from Congreve, Defoe, and Swift, and the same span separates them from Chaucer; but Chaucer’s English was nearly as foreign to them as it is to us.

Moreover, change may be quick in one aspect of a language, but slow in others. Consider the modern Scandinavian languages, which branched out and evolved from Old Norse over the past thousand years. Icelandic has by far the most conservative grammar of the family, but its phonology is the most wildly innovative - if Erik the Red were to return to present-day Reykjavík, he would be able to read the newspapers without much difficulty, but he would scarcely be able to understand the same articles read aloud. In Copenhagen, he would have the opposite problem; Danish pronunciation is far closer to Old Norse, but its grammar has diverged much more.

In absolute terms, linguistic change is unpredictable - we cannot tell what changes will occur, or how rapidly they will catch on. However, we can identify factors that make a language more or less susceptible to far-reaching and rapid change.

Language contact is the factor most likely to contribute to linguistic change - most obviously, through the borrowing of new words. English has borrowed so many words, primarily from Norse, French, Latin, and Greek, that a large majority of its lexicon is now foreign in origin, though the most common English words are nearly all native. While English tends to simply adopt new words as they are, however odd they sound, that’s not the only way to borrow new vocabulary - complex words may be borrowed through the process of calquing, or loan-translation, in which the parts of the word are translated separately and then reassembled. "Superman" is a calque of the German "Übermensch." Through borrowing words, languages may borrow new sounds - if speakers hear enough loanwords containing a foreign sound, it may come to seem less foreign, and a new generation may grow up thinking of it as just one more of their own language’s sounds, and use it freely when coining new words. Phonological rules are also susceptible to borrowing - if large populations begin to speak with the accent of a neighboring language, aspects of that accent can spread and eventually become the default.

But borrowing is not the only way that language contact fosters change. Consider Old English and Old Norse, which were mostly mutually comprehensible - there is no mention of translators in the exchange of threats and boasts in "The Battle of Maldon" - but differed in their inflectional endings. Language contact in the Danelaw counties of northern England seems to have contributed to the loss of many of those endings, and the use of a much stricter word order to convey subject and object. Contact with an unrelated language may also put pressure on a language’s grammar - languages which are often learned by adults as second languages tend to have more regular grammars than languages which are used almost entirely by native speakers.

The most extreme language contact situation comes about when adult speakers of several languages are thrown into contact and devise a pidgin to communicate - a core vocabulary which each speaker uses according to her own language’s grammatical rules. What common grammatical conventions do arise tend to operate using whole words - for example, duplicating a word as an intensifier or a plural marker. Such grammatical kludges will be codified as hard-and-fast rules, however, if a generation of children grows up speaking the pidgin as a native tongue; they will convert it in one generation to a creole, a fully-fledged language with a regular and consistent grammar.

Creoles have arisen in several parts of the world with English as the source of most of their vocabulary - what is known as the lexifier. Tok Pisin, spoken in Papua New Guinea, is the best-known such English-based creole; others include Miskito in Nicaragua and Honduras and the creoles of Grenada and Carriacou. Though many differ wildly from English in grammar and pronunciation, the shared vocabulary items can make them at least partially comprehensible to English speakers. Designing an English-lexifier creole would be nearly as major a conlanging endeavor as designing a language from scratch, but the result could easily be shown as completely incomprehensible, or largely transparent, or anywhere in between - an advantage if you want to show a character learning a language, especially if you need your readers to pick up the language faster than the POV character.

Other things besides language contact promote linguistic change, of course. Large populations permit many dialects to develop; these dialects sometimes diverge so widely as to be mutually incomprehensible (in this way are new languages born) but more often function as linguistic R&D labs - sources of innovations that may spread into other dialects, or into the standard language, if there is a standard. Social mobility also promotes change, as people strive to sound different than their neighbors and their parents.

A FINAL WORD ON DETERMINISM It can be a useful shorthand to provide your future society with a range of highly specialized terms for its greatest preoccupations - but this shorthand tells us only how people in that society spend their days and make their living; it tells us little or nothing about how they think. All specialists, whatever their specialties, create and use jargon; English-speaking skiers have more words for snow than Eskimos.

Conversely, to tell us that a people have no word for a thing is not to tell us that they do not have the thing; still less, to say that the thing is inconceivable to them. The content of one’s thoughts is not determined by the extent of one’s mental dictionary. The Sapir-Whorf hypothesis is nearly dead in linguistic circles from lack of experimental corroboration, and linguistic determinism, as a science-fictional idea, is on a par with phlogiston: not to be used without a healthy appreciation of its wrongness.

That language and thought are separate things, that thought does not require language, can be difficult for writers to conceive - after all, language occupies our thoughts to such an extent as to drive out nearly all other concerns. The historical popularity of linguistic determinism in SF is easy to understand: it is one of the only ideas to come out of linguistics in the last century for which lay writers could easily derive thought-experiments. And the impulse to put language under a microscope, to center a story on it - what could be more natural for writers?

But that impulse overlooks the forest for the trees - language cannot be anything but the center of any story, and every word on the page is already under the microscope of the readers’ scrutiny. To write well and engagingly about language in an SF setting requires only that one write good and engaging SF. If you’ve done your homework - if you know what languages your characters speak, how they developed, whether they’re mutually comprehensible, and have a few pages of notes that may never get into the final draft - then the readers will only notice what you want them to notice; in this respect, getting the linguistics right is no different from getting the physics right. And getting the linguistics right takes you one step closer to getting the language right: and that, of course, has always been the goal.