So, Your Utopia Needs a Language...
By Tristan Davenport
24 October 2005
"The limits of my language stand for the limits of my world."
Ludwig Wittgenstein was overstating his case when he wrote the famous sentence above, but that has not prevented science fiction writers from taking this concept and running with it. Ever since George Orwell, SF authors have been inventing languages to explain or illustrate the alien psychologies of their characters. If Eskimos have fifty or a hundred words for snow, they reason, how many words will future man have for spaceships? If Klingons love only war, will they have even a single word for "love"? Such questions make sense to us because we have swallowed the extreme notion that our language limits, even determines, our way of thinking. This concept is called linguistic determinism. It comes in a strong form—Wittgenstein's absolutism—which science has largely rejected, and a weak form, which has won some acceptance. But it is the strong form that has quietly taken residence in SF, to the point that today no fictional society is complete without a matching language.
This essay is about artificial, or constructed, human languages—as seen both in real life and in fiction. First, I'll discuss some of the more famous examples of real and imagined constructed languages. I'll move on to examine how natural languages vary, and how some of these distinctions—which may at first seem bizarre to English speakers—can be built into a language of your own to add color and depth to your speculative fiction. Finally, I will discuss the Sapir-Whorf hypothesis, the form of linguistic determinism that has probably had the greatest influence on SF.
Constructed Language in SF
In the realm of fiction, the first constructed language that comes to mind is Newspeak, the language of Oceania in George Orwell's 1984. Newspeak is the poster child for linguistic determinism. As Orwell notes in the novel's appendix, it is the only language that actually shrinks with the passing of time: each year the official dictionary is thinner. The philosophy behind Newspeak is that if a concept cannot be verbalized, it cannot be conceived of at all. How can a policy be wrong, for example, if there is no word for "wrong"? Newspeak relies on the thesis that thought consists of words. Without a word for a thing, we cannot think it. Newspeak shrinks to leave no room for dissent.
A counterpoint to Newspeak exists in Gene Wolfe's The Book of the New Sun. The tyranny in that book is the government of the northern country of Ascia. Like Orwell's Big Brother, the rulers of Ascia control their subjects through a restrictive language, intended to restrict thought as well as expression. In this case, the restriction is not in the number of words used, but in the number of utterances considered acceptable. A normal Ascian has not spoken a natural language (a language of words and rules allowing for infinite possible sentences) since early childhood. Instead, Ascians are trained to speak from a list of proverbs and accepted sayings—something like quotations from Chairman Mao's Little Red Book. Original utterances are forbidden, and the only means of expression are quotations from the officially sanctioned text. But Wolfe is more optimistic than Orwell. Language does not determine thought after all: the Ascians subvert the official grammar. By using the figurative and metaphorical meanings of the quotations, they are able to express nonsanctioned ideas.
On the brighter side of things, we have Ursula K. Le Guin's Anarresti. This is the language used by the anarchist colonists in her novel The Dispossessed. A bit of history: to escape the injustices of the gung-ho capitalist planet Urras, a group of anarchists settle its large moon, Anarres. In contrast to Newspeak, the tongue they invent for themselves is designed for equality, freedom of thought, and harmony. The language encodes anarchist principles, dispensing with phrases that might suggest exploitation. Possession, for example, is not directly expressed in correct Anarresti. One says "the pen I use" rather than "my pen." Likewise, Anarresti does not distinguish between passive and active sexual roles. The English sentence "I rogered Cindy" would be expressed in Anarresti as "Cindy and I rogered together." Sexual violence, Le Guin implies, does not happen on Anarres, presumably because the language gives nobody the idea.
In Samuel R. Delany's novel Babel-17, the language that gives the book its title seems to occupy some zenith of logic and expressive power. It is, as one character describes it, "the most analytically exact language imaginable." Babel-17 is so structured that the word for any common object conveys meaningful information about the object's form, function, or makeup. In one scene, the heroine awakes to find herself tied to a bulkhead. Examining the net that holds her in place, she names it in Babel-17. The word for the net contains information about its structure, which allows her to discover its weak point and free herself.
Like 1984, Babel-17 espouses a strong form of linguistic determinism. In the novel's universe, thought is language. The conceit of the novel is that a language can be designed as a weapon. Babel-17 encodes in its very structure a psychopathic personality that quickly takes over the speaker's mind. Babel-17 is able to do this, presumably, by virtue of its great efficiency as a medium for ideas—such great efficiency that just learning the tongue bestows on the heroine a commensurate efficiency of thought. When she thinks in Babel-17, time itself seems to slow, connections once subtle become obvious, even her reflexes speed up. Efficiency, it seems, is everything.
A Perfect Language
Modern linguists agree that the notion of one language being more efficient or more expressive than another is pretty much hokum. A few examples will show why:
Language | Word | Literal meaning | |
| Slovak | počitač | "counter, thing that counts" | |
| Inuktitut | qarasaujaq | "thing that works like a brain" | |
| Icelandic | tölva | "number-seeress" |
In their respective languages, the three words refer to the everyday object we call a computer. The Slovak word gets at this meaning in a very literal way, by describing the object's function. The Inuktitut (Inuit) word captures the meaning by comparing the computer to a more familiar object, the human brain. Icelandic opts for a third tactic, metaphor.
Not only would a "perfect" language convey useful information with the clarity that Slovak does, it would also be obvious to the learner like Inuktitut, not to mention poetically rich like Icelandic. Clearly, a language cannot do all three at once. Nor does it need to, in order to help its speaker survive and thrive in a human society. That, after all, is what language evolved to do.
Despite the obvious sufficiency of natural language, the notion of a perfect, artificial language persists. The recent history of philosophy is littered with the ruins of limpidly structured, impeccably logical tongues. Few of them found any speakers. Constructed languages that aim for ease of learning have succeeded somewhat better. Esperanto, invented in the late 19th century, is the most famous example.
The vision of L. L. Zamenhof, Esperanto's creator, was to bridge cultures with a grammatically regular, ethnically neutral language. Perhaps it's ironic that Esperanto is explicitly based on European models. Esperanto succeeded in being easy to learn, and has become by far the most successful constructed language. Estimates of the number of speakers vary widely—from 100,000 to nearly two million.
If Esperanto is a relative blockbuster, Basic English went straight to video.
The project of Basic English was a noble one: create a simplified, compact version of English that any foreigner could easily learn and use. Basic English followed standard grammatical rules but employed a vocabulary of only 850 words, eighteen of which were verbs. A brief sample:
Basic English | English |
| I'd like the hot liquid with a sweet cover, made with the plant from the earth that makes eyewash. | I'll have the French onion soup. |
Like the Clapper and the Pet Rock, Basic English initially met with enthusiasm. Winston Churchill supported the idiom, [1] and H. G. Wells predicted that it would become the world language. But for reasons captured in the above quotation, it failed to catch on.
The last constructed language I'll consider is perhaps the most famous in the genre community: Klingon. Cooked up by a linguist at UC Berkeley (where else?), the Klingon tongue was commissioned by Paramount Studios to add color to the Star Trek universe. Linguistically, it's notable for doing everything that familiar, Indo-European languages do not. A sentence's subject, for example, comes after the object in Klingon—a word order that's exceedingly rare on Earth. Likewise, Klingon employs a grammatical gender system that may seem unusual to English speakers: rather than classifying nouns as male or female, animate or inanimate—as most European tongues do—Klingon classifies nouns as body parts, sentient beings, or neither. These are the three genders of Klingon. These features are built into Klingon not to "explain" the Klingon view of the universe per se, but simply to add exotic flavor.
Inventing a Language
So, that is the first thing a constructed language can add to your speculative fiction: atmosphere. In such cases you need not plan very much of the language. All that's needed is a general idea of how the language is going to sound—which phonemes (basic sounds) the language uses and how it tends to combine them. What makes your language memorable in the way that "La bave du crapaud n'atteint pas la blanche colombe" is instantly recognizable as French? Do not strain too hard for originality. Keep in mind that the human mouth works best with vowels surrounded by consonants. Words like "J'k'Fll" may sound exotic to you, but to an editor they mark a novice typist who has yet to master the home row.
In imagining how your language sounds, it pays to keep certain linguistic universals in mind. Certain phonemes are simply more common than others. Almost every human language, for example, has the consonants /p/, /t/, /k/, /m/, and /n/. The most common vowels are the five used in Spanish: /a/ (as in "hot"), /e/, /i/ (as in "beet"), /o/, and /u/ (as in "food"). Some examples of less common phonemes are the /th/ sounds found in "ether" and "either" and the /q/ sound (like /k/, but made against the uvula) found in Arabic and in many Native American languages. [2]
So much for sounds—what about grammar? Here are some of the most important things you'll have to consider. The first is morphological typology. This concept is not as scary as it sounds. What it means is that different languages have different ways of putting words together from smaller pieces. Any word, or part of a word, that can be said to have a meaning of its own is called a morpheme.
Consider the English word reopening: it consists of three morphemes. The prefix re- (meaning "again"), the verb stem open, and the suffix -ing (meaning "this action is not finished"). English is a fusional language: it makes words by starting with a root morpheme and adding prefixes and/or suffixes to it.
Consider now a word from Turkish: Gidebileceksin "You will be able to go."
| Gid- | "go" |
| -ebil- | "able/can" |
| -ecek- | [future tense] |
| -sin | [2nd person singular (you)] |
Turkish is agglutinative. It operates by gluing morphemes together. Words, especially verbs, tend to contain many morphemes. Words in fusional languages such as English consist of one morpheme or only a few, and they are often combined in irregular ways, such as get → got. Agglutinative languages tend to create words in a much more regular fashion.
Some languages are so adept at combining morphemes that a complete statement or question can be expressed in a single word. Such tongues are polysynthetic. Consider the word below:
Pariliarumaniralauqsimanngittunga
The language is Inuktitut. The word means, "I never said I wanted to go to Paris." Unlike the Turkish and English examples, in which all of the morphemes in a word refer to the root morpheme or to the sentence's subject, this Inuktitut word contains a noun, two verbs, and an implied subject. Quite a lot to pack into one word! And what if that rich morphology were brought to bear on something besides Paris, such as snow? In such ways urban legends are born. [3] Like agglutinative languages, polysynthetic ones tend to follow very regular rules.
At the other end of the spectrum, there are languages like the various Chinese tongues: isolating languages. As the name implies, such languages give each morpheme a word of its own. The sentence below is in Taiwanese:
| He | jia | sanyoong | nye | hia | nie | dua | jia |
| That | (animal unit) | mountain-sheep | why | so | ? | big | (animal unit) |
The sentence means, "Why is that goat so big?" The particle jia is a common feature of east Asian languages; when counting something, a classifier like jia identifies to which category of objects the thing being counted belongs. Since the distinction is only made in limited contexts, this is not quite the same as a gender system.
Each morpheme comprises a whole word, except for the compound sanyoong, the word for "goat," which translates literally as "mountain-sheep." It's worth noting that even in this classic example of an isolating language, a two-morpheme word exists. [4] Morphological typology is a fuzzy continuum, not a rigid collection of categories. English is fusional, but less so than Spanish and Greek; Japanese agglutinates verbs but isolates nouns, etc. Linguists argue all the time about whether a given language deserves to be called agglutinative or polysynthetic, or whether the term polysynthetic should be used at all.
The point of all this information is that your constructed language can have a grammar very different from that of your mother tongue. No morphological scheme is more logical or "normal" than another. It is an accident of history that at this time, the most powerful cultures on Earth speak fusional languages. Your dominant or colonized culture need not follow suit.
Now for another aspect of grammar: word order. I talked earlier about the Klingon penchant for putting the subjects of sentences after the objects. If Bob Marley were Klingon, he would sing, "The sheriff shot I." Were he Fijian, the same line would run, "Shot the sheriff I." Indeed, for any possible ordering of the three elements—Subject, Object, and Verb (S, O, and V)—there exists a language that employs it. English is an SVO language, and Fijian is a VOS one.
This sentence-level word order ties into other aspects of a language's grammar. For instance, SOV languages (e.g., Turkish) tend to put adjectives and adverbs before the words they modify and prepositions (in this case, postpositions) after. In VSO languages like Gaelic, the opposite is true. This point may be of interest only to linguists and constructed-language hobbyists, but even if you use only a few lines of dialogue from a constructed language, it pays not to commit any easily avoided errors.
Language as a Lens to the World
Linguists, anthropologists, and perceptive travelers have long noted that different languages divide up the world in different ways. A few examples:
Speakers of the Hawai'ian language do not commonly talk in terms of the four compass directions. The common directional terms translate to "seaward" and "inland/mountainward."
When addressing someone in Japanese, the speaker must include a suffix indicating that person's relative status. This is the -san in the "Daniel-san" of the Karate Kid films. In addition, the emperor of Japan has his own pronoun, distinct from that used for other people.
The Tamil language of southern India requires suffixes similar to the Japanese ones, except that instead of status they signify the capacity for rational thought. Humans and gods are set apart grammatically from animals, abstract states, and things.
Many Papuan languages have no words for colors beyond black and white.
Linguistic determinists get into chicken-and-egg arguments all the time. Are the Japanese traditionally more sensitive to status because of their rich vocabulary for it, or did they develop that rich vocabulary because status is so important in Japanese tradition? Think about your own language. How many American slang terms exist for money? For active sexual roles? Do these terms influence your thinking? To what extent? Does linguistic relativity, the observation that languages vary in the ways they describe reality, lead automatically to linguistic determinism?
According to the Sapir-Whorf hypothesis, the answer is yes: the language a person speaks strongly influences the way he or she perceives the world. I, an American English speaker, can be expected to have a lower awareness of social status than a Japanese person does. The crucial difference is that my language does not force me to keep it in mind.
An interesting thought experiment concerns the way language expresses the two genders. Imagine a country in which the language all but disregards the male/female difference. In country A, separate words exist for man and woman, perhaps for brother and sister—only the bare minimum—but a single word contains both he and she, likewise waiter/waitress, chairman/chairwoman, and all the other sexually dimorphic pairs found in English. Except for the basic terms like man/woman, the language simply ignores human gender. Now imagine the neighboring country B, in which the language explicitly and mandatorily encodes gender data. Not just he/she is present, but also paired terms for every word that could be used to describe a person: "pilot/pilotess," "pacifist/pacifistess." Adjectives, too, are inflected for the gender of the person they describe.
So which country would you expect to be more egalitarian, country A with the genderless language, or country B with its rich gender structure? Sapir-Whorf, entangling language structure with worldview, would seem to predict country A. The actual answer is that it goes both ways, independently of language. Our pair of countries can be Finland and Sweden, both among the best places in the world for a woman to live. Genderless, agglutinative Finnish is as different as can be from the gender-inflecting, fusional Swedish language. Yet both countries are at the world's forefront in women's health, education, and political involvement.
We could also select a very different pair of countries. The Persian language is quite genderless, yet it is arguably even more unfortunate to be born female in Iran than in Saudi Arabia, where the language uniformly encodes gender distinctions. In all four cases, language is far less important than economics and local history in determining people's attitudes toward gender.
Such failed predictions are the reason that only the weak forms of linguistic determinism are still debated. Speakers of the two-color Papuan tongues have more difficulty than English speakers in remembering colored objects. Yet they seem to perceive color (perceive being a key word in the Sapir-Whorf hypothesis and other strong forms of linguistic determinism) without regard to their so-called handicap: concepts like "red" and "blue" make just as much sense to these Papuans as they do to speakers of other languages. By the same token, English speakers have proven themselves just as capable as any people of functioning in hierarchical systems, such as monarchies, corporations, and school yards. Sapir-Whorf may go the way of Carl Jung—more a metaphor than a rigorous theory, of more use to artists than to researchers.
So how can you use it? Samuel Delany, who has a great many things to say about language, provides some excellent examples. His SF novel Stars in My Pocket like Grains of Sand is narrated almost entirely in a language that distinguishes two genders: "he," a person to whom one is sexually attracted; and "she," everyone else.
The Left Hand of Darkness by Ursula K. Le Guin describes a language in which three grammatical genders are recognized: hermaphrodite, temporarily female, and temporarily male.
These authors are right that male/female is just the beginning of gender—certainly that's true as far as language is concerned. "Gender" in linguistics refers to nothing more than a system of classifying nouns. Some languages, like Turkish, have no genders; others have as many as forty. Recall the Klingon gender system, or consider the Wari one. Wari, spoken in the Amazon rainforest, assigns every noun to one of two genders, edible or inedible. Members of enemy tribes are in the edible category.
Delany's Babel-17 takes the Sapir-Whorf hypothesis to perhaps its logical extreme: the language Babel-17 does not contain terms for "I" and "you." One major character, who happens to speak Babel-17, is completely unable to form a concept of self, because the language lacks a term for "I."
Delany's position may have been trendy in the sixties, but research in cognitive science now strongly indicates otherwise. [5] That one can have a thought but be unable to verbalize it is proof in a nutshell that thought is something more than language. Linguistic "oddities" such as the Wari genders and the Papuan two-color system may be highly interesting, but it is unlikely that they unalterably limit a speaker's understanding. Humans are flexible, able to think beyond the confines of language, which is but one of our many tools for survival.
In Your Own Words
When you use constructed languages in your own fiction, consider the following questions as jumping-off points. What other meaning distinctions are possible? How do your characters relate to the world, and how (and to what degree) is it expressed through language? Is their language a natural one, or did someone set out with a conscious plan to create it? Matters of aesthetics are also important. What are the rhythms and phonemes of your language? Does it use tones like Chinese, a pitch accent, nasal vowels?
In many cases, it is easier to borrow than to invent. A good strategy used by J. R. R. Tolkien, himself an expert on Old Icelandic and other northern European languages, is to find an Earth culture that resembles your fictional culture. Research the language, get an idea of its spelling/writing system, grammar, and phoneme inventory, and model your fictional language on that. Tolkien's elves were associated in his mind with the Celts of the British Isles. His constructed language Sindarin resembles Welsh. Your language can sound like the language of the Hopi, the Malays, the Vikings, the Zulu. Whatever mixture of invention and borrowing you decide to use, thinking about your story in terms of language can open up new vistas of meaning and metaphor.
Notes:
[1] Winston Churchill is said to have reconsidered his position on Basic English when he learned that "blood, toil, tears, and sweat" would be rendered "blood, hard work, eye wash, and body water."
[2] The phonetician Peter Ladefoged has compiled an excellent web resource for fans of exotic speech sounds: Vowels and Consonants.
[3] By one recent count, Inuktitut has only three morphemes that signify different kinds of snow. When Douglas Adams wrote that Inuktitut has a word for "The snow back in my day that was much better than this modern snow," he may not have been joking. It's possible that that entire phrase could be translated as a single Inuktitut word.
[4] Although sanyoong is written with two characters, each one allegedly a whole word, it is pronounced by native speakers as a single compound word.
[5] Steven Pinker addresses this idea and a host of others in his accessible books How the Mind Works and The Blank Slate.
(Comments on this piece | Articles Forum | Main Forum Index | Forum Login)
