How to Invent an Alien Language? A Linguistic Perspective
Invented languages spoken by races in imaginary worlds can add credibility to a story, making both the world and the characters feel more real. An author of a science fiction or fantasy novel involving different countries or planets may choose whether to make everybody speak the same language or try to create separate languages. As pointed out above, the latter option can increase credibility, but it comes with its own complications.
Incomprehensible sentences make the text quite heavy-handed and an intermediate variant should not be neglected. A character from a different culture (whether an alien or an elf) could talk the same language as the protagonist (i.e. the language in which the book is written) but with certain systematic mistakes. These mistakes would give a clue about his or her own mother tongue without making the dialogues unreadable or even unclear. Credibility will not come at the expense of comprehensibility.
However, in order for the character’s speech to be believable, the mistakes have to be systematic, or rule-governed. Otherwise, he will just give an impression of an illiterate guy, or of some funny absent-minded stranger, neither of which may be what the author intends. If a whole new language with its own lexicon is invented, the existence of rules and patterns is even more essential. There is no language without rules, no matter how far from Earth it is spoken. But how does one create such a language?
There are two ways to go. One could try to invent a communication system whose rules differ radically from what is actually found in languages of the world. This is indeed a good way to reflect a dramatic contrast in culture and mentality. But, interestingly enough, authors who choose this direction quite often end up “inventing” linguistic characteristics which do, in fact, exist in some real languages (albeit rarer ones). One reason has to do with the fact that world languages are quite diverse.
The other reason is, in some sense, exactly the opposite. There is considerable evidence that our linguistic capacity has an innate component. Of course, the knowledge of English, Russian or Chinese is not inborn, but we are genetically predisposed to acquire language. Universal restrictions, deeply embedded in our minds, make it extremely difficult to even think of rules that would be totally impossible for any human language.
Of course, coming up with such a language would be great for the purposes of showing how different the race in question is from humans. But it is also possible to use what we know about human languages as they are for the same purpose. Specifically, we can look for those rules that exist in world languages but are very rare, and use such rules in the communication systems we invent.
Changing the order of words and phrases, while keeping the language identical to English in other respects, is a good way to make a text sound ‘alien’ but comprehensible. This strategy is illustrated well by Yoda’s speech in Star Wars.
Typological facts are helpful in this respect. Languages of the world differ in their basic word order. Consider the simple English sentence, “John loves Mary.” As you can see, the subject (John) comes first. It is then followed by the verb (loves) and, finally, the direct object (Mary) appears. This is the standard word order in English: subject-verb-object. This is why linguists call English an SVO language.
SVO pattern is very common among human languages. Another very popular pattern is SOV (John Mary loves). However, the remaining logical possible orders are much rarer. VSO (Loves John Mary) is still well represented. It is found in about nine percent of world languages, such as Irish and Biblical Hebrew. But the other patterns are found in no more than three percent of languages, if at all. These include VOS, OVS, OSV. An invented language with one of these patterns will sound different not only from English but from over ninety-five percent of languages spoken around the world!
The OVS word order, which is among the rarest ones, is characteristic of Klingon, an artificial language used in the Star Trek series. In a 2010 blog post, a linguist Asya Pereltsvaig illustrates this property by the following example. The sentence “I boarded the Enterprise” corresponds in Klingon to the order “The Enterprise boarded I.” But since Klingon is a full-fledged language, the lexicon used in this sentence will be different as well.
The syntactic rules of Yoda’s language are somewhat more difficult to formulate. As linguists have noticed, some of his sentences are totally identical to English ones (e.g. “You must unlearn what you have learnt”), whereas others use different word order rules. Often, the OSV order can be observed, which, too, is one of the rarest typologically: “Much to learn, you still have.” This tendency is noted, for instance, by Geoffrey K. Pullum in a 2005 blog post.
Yet another element one may consider is the relative order in such phrases as “on the table” or “after lunch.” While in English, on and after are prepositions, i.e. they appear before the noun, some languages use postpositions instead. So, in these languages we would rather get “the table on” and “lunch after.”
Postpositions, found e.g. in Hindi, Finnish, and Turkish, are not as rare as the OVS word order, but their use would make a language sound considerably different from English, Spanish, Russian, and many other tongues.
Let us now move from sentence structure to the internal structure of a word. We often add prefixes and suffixes in order to create new words or new word forms.
For instance, -ment in management is a suffix (since it follows the stem), while re- in reread is a prefix (it precedes the stem). Both suffixes and prefixes are quite common cross-linguistically. However, infixes, which are inserted inside a stem, are virtually never found in Standard English and are, in general, rare in Indo-European languages. They are found in Austronesian languages, however.
For instance, consider the following pair of words in Tagalog: gradwet ‘graduate’ and grumadwet ‘to graduate’. (The similarity in sound to the English graduate is non-accidental.) In the second form, the infix -um- is inserted before the first vowel of the stem. Occasional use of infixes will make your language sound unusual.
Yet another property that distinguishes world languages from one another is case. Consider the different forms of the pronoun in the following two sentences:
He loves Mary.
Mary loves him.
We find he in the subject position but him in the position of the object. We also have the form his when the relation of possession is expressed. This is what case is about. Roughly, the same noun appears in different forms depending on the position it occupies, or on its grammatical function. Subjects, direct objects, indirect objects, nouns that follow a preposition, etc. may appear in different forms, depending on the rules of the individual language.
In Modern English (unlike Old English), case differences are found mainly on pronouns. But in many languages, the phenomenon is extended to other nouns, such as John, table, or freedom. Languages differ enormously in the number of case distinction they make. Some have just two cases, roughly, he versus him (e.g. Romanian), others three (Greek), yet others four (some Germanic languages), and so on. Finno-Ugtic languages exhibit more than ten case-forms, and some Daghestanian languages have been argued to have more than forty cases! The Klingon alien language employs this feature as well.
Even though the phenomenon is far from being rare in reality, adding short case suffixes to nouns fulfilling different functions will create a foreign flavor for the ear of an English speaker.
Consider the difference between such nouns as cow and cattle. The former can easily be pluralized: cow-cows; the latter normally isn’t. (Cattles sounds pretty bad.) Also, the former can directly combine with number words, so that we can count: one cow, two cows, three cows, etc. In contrast, in order to count cattle, we need the help of a “counter word,” specifically, head. Thus, instead of the ungrammatical five cattles we get five head of cattle. What we see here is the difference between count nouns (like cow) and mass nouns (like cattle).
Why should this all matter when we talk about rare linguistic phenomena? After all, aren’t all languages like English in this respect? Well, it turns out that they aren’t. There exist languages in which all nouns behave as if they were mass. Chinese and Japanese are classical examples.
In order to combine a noun with a number word in such a language, you always need some counter word, called a classifier (like head in five head of cattle or piece in five pieces of furniture). And so, we get things like three men of student (instead of three students), four things of book (instead of four books), and so on.
Crucially, as you can see in the above examples, the choice of the classifier depends on what is being counted, e.g. man for human nouns and thing for inanimate objects. Different languages have different classifier inventories and thus use them to mark different distinctions. A distinction may be based on animacy (human, animal, inanimate), on shape (flat objects, round objects, cylindrical objects, etc.), or on other properties (for instance, a special classifier for mechanical objects).
One advantage of classifiers is that when you use them in an invented language you can decide which distinctions are going to be particularly basic and important for its speakers. Do they care about animacy, about shape, about the contrast between organic and non-organic, or about something totally different? How do they classify entities into groups? By making this choice you can, through the use of language, tell the reader a lot about the environment and/or the psychology of the creatures. Moreover, you can demonstrate their attitude towards a particular thing or kind. For example, to humans. If your aliens use for humans the same classifier as for themselves, this suggests that they treat us as equals. But if they use a different classifier, and it’s the one that is normally found with animal nouns, we can draw a different conclusion. And in case humans receive the same counter word with chicken, pork, burgers, and sausages, it’s time for them to escape the planet, since apparently, they are treated as edible objects!
Going back to Klingon, we find a strongly related property in this language: different plural markers are used depending on the meaning of the noun. The crucial division is into individuals who are able to use language, body parts, and entities that do not belong to any of these two categories.
If an author writes a whole text spoken in an elfish, dragonish, or alien language, creating totally new words with totally new roots is almost inevitable. But what if you follow the different strategy and make the characters speak English (or whichever language the book is written in) with a number of elements alluding to their true mother tongue? Would it be useful to add words here and there that do not exist in reality and are thus unfamiliar to the reader?
This could be useful, as long as several rules are observed. First, do not exaggerate. A high number of such words would make the text unreadable. Second, it is preferable to create such new words for new concepts. For instance, do not invent a new string of sounds in order to convey the meaning table. If the character speaks fluent English (or whichever language English represents in the book), why would she suddenly substitute table by her native word? This is not totally impossible, but not particularly likely either. But it would be great to have new words standing for those concepts which are characteristic specifically for the culture and the biological environment of the alien and which lack appropriate counterparts in English. This way, the biological or cultural difference is emphasized by language use.
The invented language may also make more distinctions than English does in a field that plays a particularly important role in the life of the relevant race. For example, it is well-known that Eskimo languages have more words for snow than English does. Actually, linguists have suggested that the difference between Eskimo and English in this respect has been exaggerated.
English does make lexical distinctions between such notions as snow, avalanche, blizzard, frost, flurry, sleet, (snow)flake, etc. Still, a contrast does exist. To illustrate, Eskimo distinguishes between snow fallen on the ground, drifting snow and fallen snow floating on the water. An invented language could make analogous distinctions within a different field. That will not only create the “foreignness” feeling for the reader but will also help lift the veil on the unfamiliar culture.
Our knowledge about real languages may be quite helpful in the process of inventing or even understanding as a reader, a non-existing one. If worked out appropriately, elements of such a language may serve as a powerful tool in revealing the differences between humans and other creatures, emphasizing the tension between them, and familiarizing the reader with the culture of the strangers.
Olga Kuno has completed her Ph.D. in generative linguistics at the Hebrew University of Jerusalem. She is a linguist and a writer. Her fantasy romance novels have been published by several major publishing houses in Russia. One of these novels, Half a Step Away from Love, has been recently translated to English.