Excerpts

from the work of Suzette Haden Elgin

The Linguistics & Science Fiction Sampler

Language Construction 101

You can of course construct an entire language composed of just the item "A"; linguists torment students with that kind of thing routinely. But suppose you want to construct a language that might be of some practical use for communication in your fictional universe... here's how it's done. It is not difficult.

STEP ONE: Decide whether you want a polysynthetic or agglutinating language (one where you construct meanings by assembling lots of small meaningful pieces into larger chunks, as Navajo does) or an isolating language (one like English,where words are made up of only a few meaningful pieces and often of only a single piece.) Polysynthetic or agglutinating is quicker, the way building stuff with a TinkerToy™ set is quicker; I recommend it.

STEP TWO: Choose an order for verb, subject, and object. Only six are mathematically possible; English uses the order "Subject, then Verb, then Object." Pick one.

STEP THREE: Choose the structure and assembly rules for your syllables (pronounceable chunks). For example, you could decide that all syllables of your language must contain a vowel; that none can begin with more than one consonant; that all can end with either a vowel or a consonant; that no double (long) vowels or consonants are allowed; and that no more than twelve syllables may be in a single word.

STEP FOUR: Choose a set of phonemes (that is, chunks of sound that change meaning.) For English, the fact that we understand "bat" and "sat" as two different words proves that the sounds of "b" and "s" in those words are two different phonemes. Hawaiian has eleven phonemes, English has about thirty-five, seventy is roughly the upper limit, and all human languages choose from the same set. (You could pick sounds no human language uses, of course, if you’re constructing a language for ETs, but you couldn’t be sure that your human readers would be able to pronounce it in their heads as they read; it’s not wise to annoy your readers that way.) Suppose we pick these twelve: /b/, /g/, /s/, /1/, /m/, /h/, /w/, /a/, /e/, /i/, /o/, /u/.

STEP FIVE: Set up an inventory of syllables that your rules will allow, either by hand or by computer. Like this....

a, e, i, o, u, ba, be, bi, bo, bu, bab, beb, bib, bob, bub, baba, bebe, bibi, bobo, bubu, bubab. . .

...and so on, till you've listed as many as you feel you need to get started with. By hand — which is how I did Láadan — this is tedious; a computer will whip the whole list out for you in a flash. The length of the flash is determined by what you did in Steps Three and Four. Obviously, if you only allow three-syllable words in your language and you have only seven phonemes, the list will be shorter than if you’re using ten syllables and twenty phonemes.

STEP SIX: Decide how you want to handle your basic grammar markings. That is: How will you mark something as plural? As past (or other) tense? As completed or still going on? And how will you indicate whether something is the subject, the object, the possessive, etc.? Write the rules you need to do these things. Suppose you decide to mark these basics by adding syllables to your words. If you do that, you’ll have rules like these: "Adding 'ba' at the end of a word makes it plural." "Adding 'ga' at the beginning of a verb makes it refer to past time." "Adding 'fa' at the end of a word marks it as the subject of your verb." And so on.

STEP SEVEN: Start assigning meanings to your listed syllables, for your core vocabulary. That is, for words like "house, woman, child, man, tree, fire, make, eat, drink"... and words that are totally invented as well, if you need them because they’re as basic to your fictional culture as "fire" is to human culture.

STEP EIGHT. Make your basic decisions about syntax. That is: How will you indicate that a sentence is negative, or is a question, or is a command? And-- very important — how will you combine two or more sentences into a single bigger sentence? Once again you could do this by using syllables. Like ... "Adding 'fo' to the last word in a sentence indicates that it is a question." "Adding 'wa'at the end of the first word of a sentence indicates that it is embedded inside a larger sentence, as 'Mary is tired' is embedded inside 'I know Mary is tired' in English." And so on. No human language does these things by repeating words, but for an ET language you could decide to do exactly that. You could have a rule that said "A sentence in which every word is repeated twice is a question." "A sentence in which every word is repeated three times is a command." And so on. Human beings would find that cumbersome, but your ETs might not; that’s up to you.

STEP NINE. Take some simple text... a short folktale is a good choice... and start translating it into your language. This serves as a diagnostic probe to let you know what you need to add or change. For the Láadan language, which was constructed to express the perceptions of women, I began by translating the Twenty-Third Psalm, because the King James Bible is one of the most masculine-perception-expressing books I know of and that psalm is the right size.

And there you are; this is how it's done. When you get through with these steps you will have a usable language, meeting all the specifications for a usable language. That's just the beginning, of course. Turning it into a living language would require native speakers. Using it to write great literature would require many years of additional elaboration, plus writing talent. Using it as part of a story or novel would require the construction of a culture to go with it. Turning it into a best-selling grammar and/or audio program would require a powerful media unit like the group behind the Klingon materials. Nevertheless, you would have a whole language, constructed by you, to your personal specifications, to use for whatever purposes you like.

Book Review

Aliens and Linguists: Language Study and Science Fiction, by Walter E. Meyers. University of Georgia Press 1980. ISBN 0-8203-0487-5. Hardcover; 257 pages. Notes, bibliography, and index.

I personally would be fascinated by any book that had as its subject the interface between science fiction and linguistics. Even if it were a compendium of errors and distortions, which this book most definitely is not. But I'm convinced that Aliens and Linguists would hold my attention even if I had no interest in either of its paired topics.

On page one of the first chapter (titled " 'The ‘Godlike Science’ and Science Fiction"!) Meyers tells us that "the first thing to be noted is the striking contrast between the wealth of language problems in science fiction and the relative poverty of linguistic explanation." He points out that any sf story extending over time involves language change, and devotes chapter 2 to that subject. Chapter 3 discusses Alien beings and the 'differentness' of Alien communication systems, and (together with chapter 4) points out that "the first aliens we need to deal with, indeed that we are already trying to speak to, are located not in a distant star system but right here on earth." (Page 10; he's talking about the apes and the dolphins.) Chapter 5 moves on to non-Terran intelligent Aliens, and discusses how the linguistic search for the universals of human language relates to the variety of communication systems authors have provided for their ETs. Chapter 6 expands on this question, showing how crucial the concepts of linguistic theory are for all first-contact stories. In each of these chapters Meyers discusses science fiction stories in which somebody has to learn a new language, to illustrate and clarify the points he's making.

Chapter 7 looks at a number of ways writers have tried to make language-learning less difficult; Chapter 8 looks at their attempts to find ways to avoid language problems, including the automatic translator. Chapter 9 takes up language-related questions of telepathy. Chapters 10 and 11 discuss some sf classics whose major focus is language and use them to present "A History of Linguistics in Science Fiction." Finally, chapter 12 examines "some of the things that linguistics and science fiction together can tell us about society." (Page 10.) There are extensive notes for all the chapters, an excellent bibliography, and a useful index.

In chapter 1 Meyers takes up a situation that has had a certain amount of attention: Sf prides itself on its scientific accuracy, but uses for the most part abominably ignorant linguistics. He also points out a closely-linked situation that I think no one else has mentioned — that the sf writer can ordinarily assume that the audience written for is well acquainted with at least the basic concepts of the science treated in the book, so that blatant errors will be noticed. He's quite right. The sf writer who misplaces a star in the heavens or miscalculates a parabola will get baskets of letters from readers, all gleefully pointing out the scientific error. But errors in linguistic science, even at the most elementary level, slide right by. Public ignorance of linguistics (which is in my opinion pri-marily the linguists' fault) makes this possible. Meyers states all this explicitly and tackles it head on. Each time he discusses the treatment sf writers have given some linguistic concept, he begins with a lucid and careful explanation of that concept for the nonlinguist reader.

One of the most glaring areas of linguistic nonscience in sf, Meyers notes accurately, is the treatment of historical linguistics. Most sf writers deal with language change over time by adding a new slang term or two plus a sentence like "Of course they were all speaking the English of 3012" and letting it go at that. Characters from the 20th century U.S. who turn up in 30th century America rarely have communication problems due to language change, though they may get into diffi-culty with changes in widgets and gizmos; the same is true for stories sending English speakers of today back through time. Writers do have problems here, even when they know the science well. When I wrote the Ozark Trilogy, for example, I wrote about a fictional population that left the American Ozarks in the year 2012, speaking Ozark English, and then spent a thousand years on another planet in complete isolation from any other English speakers. I'm trained to extrapolate from today's English through predictable changes — I know how to do that — but the result would be speech no reader of sf today could understand. I can't sell books that way, just as I couldn't sell books whose characters spoke authentic Old English. So my Ozarkers do trivial things to show language change over time like spelling "dollars" as "dollers" (a change I had to fight fiercely for with copy editors) and using "tadling" to mean "small child." I shored up this unscientific lack of change as best I could, specifying a system of public education in which every child, from infancy on, watches endless hours of archival videos made by the adults in their early years on Planet Ozark. Not a perfect solution, but the best one I could manage — and it did at least acknowledge that languages ordinarily do change drastically over time.

The point here isn't that readers are being asked to accept a literary "convention," the way they accept talking animals in fables, and are agreeing to do that for the sake of getting on with the story. That would be perfectly all right. The point is that neither reader nor writer, so far as anyone can tell, is aware that there's a problem. Those few writers who've altered their language enough to annoy readers have still produced only unscientifically trivial change.

Meyers is generous and thorough with capsule descriptions of sf languages. For example, Mechanese, a language (in Pohl and Williamson's Starchild) which a giant computer has invented to rule the world. Mechanese uses only one syllable per sentence — like gorb, which means "My transistors are tired" — and this of course demands a huge syllable inventory. "To achieve the requisite number, Mechanese uses tones as phonemes; its speakers must absolutely distinguish fifty separate musical pitches. The pitch discrimination of Mechanese acolytes must be perfect (they are helped somewhat by strings of 'tonal beads' which they carry and use like pitch pipes.") (Page 49.)

Meyers does an excellent job on animal communication, providing not only the promised coverage of the strategems of sf writers but also a solid back-ground. He explains the concept of linguistic "channel" and presents the limited range of such channels sf writers have attempted so far, such as languages of color and smell and taste. (He points out the problems with those attempts; for instance, a language of smells suffers from the difficulty of erasing any one unit of communication before another one can be introduced.) He discusses the attempts to get around language learning — by hypnosis, by taking a pill, by using an "automatic translator" helmet or widget. As he says on page 117, summing it all up, "Writers of science fiction seldom spare their characters: they may slam their heroes' ships into planets or send their heroines to kill tigers with knives; they may freeze them into statues on Pluto or shoot them through exploding suns. Hardly any degradation or suffering is spared — with the exception of exposing them to the rigors of learning a foreign language."

The two chapters devoted to the history of linguistics in sf are superb. He starts with Tolkien, who, he says, "has not been shy in admitting that his great narrative was written to provide a source and a setting for the Elvish language he had been building since his childhood." (Page 148.) He gives us the history of the Elvish phenomenon (along with a detailed translation of a Quenya poem and a mini-glossary) and ends by noting that although there are lots of journals that specialize in the work of a single author, only Tolkien has been honored by a journal that specialized in a single fictional language. (This was true when Aliens and Linguists came out — the journal specializing in the KIingon language from Star Trek was still far in the future at that time.)

Meyers provides a wonderful discussion of the Sapir-Whorf hypothesis, stating that its central question is "Does our perception of reality constrain our language, or does our language constrain our perception of reality?" (page 160) and offering many examples from sf stories. He tells us that Whorf wrote a science fiction novel (not yet published) called The Rulers of the Universe and that "it was during the writing of the novel that Whorf began to consider the relation between language and thought, a consideration that culminated in his statement of the principle of linguistic relativity." (Page 160.) And, on page 163: "The Whorf hypothesis has a corollary: if it is true that our language determines our perception of reality, then whoever controls language controls the perception of reality as well. If language can be controlled, then would-be despots have available a subtle and efficient means of restricting thought." He goes on to discuss in detail Jack Vance's novel, The Languages of Pao, in which the Sapir-Whorf hypothesis is the entire basis for the plot.

Many contemporary linguists insist that the Sapir-Whorf hypothesis in its strong form (language controls perceptions) is nonsense, which is true, and either ignore the weak version (language constrains and structures perceptions) or call it nonsense too — which is false, and is itself nonsense. Anyone who thinks that soldiers do just as well saying "I dropped bombs on a bunch of women and children" as "I carried out an assignment in ordnance delivery" needs to come down from the ivory tower for a while.

When you read what nonlinguists write about linguistics today, you find that the general impression is that Noam Chomsky's early model of generative transformational grammar is The Very Latest Thing. That's not true, but it was more nearly true when Aliens and Linguists was written. Meyers gives us a really excellent presentation of T-grammar, using the example of Ian Watson's novel, The Embedding, to make it clear. (Steven Pinker's recent erudite best-seller, The Language Instinct, hasn't changed the public perception in any significant way, so far.)

In his concluding discussion on Utopias and Dystopias that have language as a mechanism of control, Meyers makes this important point: when writers assume that language can be used as a primary instrument for controlling people, they also assume that some magical way of preventing natural language change is available. Any elementary school teacher can tell you, correctly, that no such magic exists.

This book is a wonder; like many other wonderful books, it’s out of print. But it’s well worth hunting down through your library or your favorite used-book location; you won’t be sorry.

Copyright © 1999 by Suzette Haden Elgin