According to Chomsky, a Martian sent to Earth would conclude we speak a variety of dialects with mutually unintelligible vocabularies—but dialects, nonetheless, of a single Earthish tongue.

(Kalbfleisch 1994)

Could this be true? Could it be that—in spite of the seeming multitude of differences between English and Kwakiutl, Navajo and Japanese, Albanian and Hawaiian—every human being could accurately say, "I speak a dialect of Earthish"?

Let's assume for a moment, for purposes of discussion, that it's true; let's suppose there is in fact a language that we could (a bit more gracefully) call "Terran." What might that mean for humankind? Does it matter? Would it be important and significant, or would it only qualify as a tidbit of information to include in trivia contests?

If it's true, we clearly need the answers to a number of other questions. For example:

• Does it matter which of these Terran dialects is your own native tongue? Are some of the Terran dialects better than others—more useful, more beautiful, more powerful, or in some other way more desirable?
• If you happen to have native or near-native fluency in more than one "dialect" of Terran, are you blessed, or burdened, or neither? That is, do bilingualism and multilingualism have significant effects on human minds and lives, and if so, what are their consequences?
• Are there good reasons to work toward eliminating many of these language differences? Should we perhaps try to get rid of the hundreds of varieties now spoken only by very small populations—so that we humans could communicate with fewer of the cumbersome systems and devices now required for translating and interpreting all over the world? Or would that be a mistake?

These questions aren't trivial. We're forever hearing people say, "Oh, it's only talk!" and "Oh, we were only talking!" As if speech were some minor phenomenon of human life, something we all do but not something that deserves much of our attention. If it's "only" talk, we could save very substantial amounts of energy and time and money by stamping out most of our baroque array of ways of speaking. On the other hand, if languages are powerful and important, perhaps we should be more careful with them than we are. Perhaps we should be worried about losing any of our ways of speaking. Surely we ought to know more about what we're doing before we make decisions in this area of our lives.

We understand what it means to know one language—a single native language. Having a native language is to the definition of "a human being" what "feathered" is to the definition of "a bird." There are other creatures that are mammals, that can walk upright on two legs, that have opposable thumbs, and so on; other creatures able to communicate limited sets of messages to one another exist on this earth. But only human beings, so far as we have yet been able to determine, have native languages. Language is so important to us that the loss of the ability to use language (or the very rare failure to acquire that ability) is perceived as one of the worst of all possible catastrophes; we say of persons without language that they are "trapped" inside their bodies.

By contrast, we know astonishingly little about what it means to know more than one language, whether natively or simply fluently, and our understanding of the little we do know is limited. We don't know what adding another language means for a human being.

Opinions (even among those who don't actually disapprove of multi-lingualism) range along a broad continuum. At one extreme is the idea that multilingualism is roughly like knowing how to play two or more musical instruments instead of just one. It has a certain social usefulness, and sometimes professional usefulness—like the situation of someone who learned to play the piano as a tiny child and then learned to play the guitar as well. At the other extreme is the idea that for each language learned, the learner gains another world-view—perhaps even another "personality." Which would mean that the miracle of language is multiplied every time a new tongue is learned, or at least that the miracle grows significantly in power and degree. In between those two extremes lies every sort of variation upon them.

In this book we will consider this issue, and some related issues, from a number of different points of view. To set the stage for our investigation, let's look at half a dozen brief scenarios, set in today's United States and based on my own experience with languages and the human beings who speak them.

• Jack grinned at the relatives crowding his living room and shook his head at them in mock despair. He'd known they would find his bragging hard to believe; he'd been prepared for resistance. "I'm not kidding you!" he insisted. "I bought that Spanish computer program for Jennifer, and in two months—two months—she was rattling off Spanish like a native! And we're talking about a three-year-old kid here! I'm going to get her the French program next!"
• Thomas was determined not to lose his temper with the kids again, but he couldn't keep the sternness out of his voice as he spoke. They're not stupid! he thought. What do I have to say to get them to pay attention? "Listen to me, you two!" he said urgently. "Haven't I told you, over and over, not to let me hear you talking Tagalog? Haven't I? Hasn't your mother told you the same thing? What's the matter with you, you want to spend your whole life on welfare? You speak English, you hear me? Don't make me have to tell you again!"
• When Barbara got home from work she was in a good mood. It had been such luck that she'd still remembered enough from her years in China to be able to interpret for the patient and Dr. Gales! There could have been serious misunderstandings if she hadn't been there—and she'd saved the hospital the cost of a Chinese interpreter, too. But when she told her husband about it, he frowned at her. "I don't know if that was such a good idea," Jim said slowly. "Don't you think the other nurses may resent the way you were showing off?"
• Philip didn't like having to talk to Anna's mother about this subject; he was very worried about hurting her feelings. But it was his duty to the child, and he was determined to get it done. "Mrs. Kolyoff," he said carefully, "I'm really sorry to have to say this. Russia is a great country; I know that. It has a wonderful history of music and art and literature and scientific discovery. I'm glad that Anna can speak Russian; that's terrific. But you have to realize that it's holding her back. If you want her to go on to fourth grade, ma'am, you're going to have to convince her to speak English!"
• Greg nodded his agreement to what the others were saying about their colleague: Harold Fleming was a top-notch lawyer, no question about it! "But I keep wondering," Greg cut in. "How long are we going to be able to keep Hal in the firm? Did you know that he speaks four languages fluently? Including Japanese? He can write his own ticket, you know? He could get a top position anywhere. I just hope we can hang on to him a while, myself!"
• Janet sighed, looking at the gloomy faces around the table; it wasn't easy being a school principal in times like these. "All right then," she said wearily, "we all hate to do it, but we don't really have any choice. The only way to find the money is to get rid of the frills, and that means dropping French and Spanish. Not that foreign languages aren't a good thing, mind you! If we had unlimited funds I'm sure we'd all be solidly behind having them in our curriculum. But they're not important the way football and math are important, and—that's all there is to it. They've got to go!"

These six incidents are typical of the confusion in our attitudes about the power and importance of languages. Is it a good or a bad idea for people in this country to have command of two languages, or of more than two languages? What does it mean, for them and for others and for society? Is the learning of languages a luxury that should only be pursued by those rich enough to afford it on their own, like horseback riding, or should it be a goal of our public educational system, paid for with our hard-earned tax dollars? Should we have an international language, or is that a silly (or perhaps dangerous) idea? Do languages have the power to shape our lives as individuals and as a nation? If they do, what are the implications of the decisions we make about which languages we favor and which we ignore? Obviously, we haven't yet made up our minds.

Attitudes about languages have changed over the past three decades, sometimes for the better. I don't believe that any employer today would tell me, as did a man who was my boss in the 1950s, that he'd fire me if he ever saw me studying a Russian grammar during my lunch hour again. "What on earth will our customers think if they see you doing that?" he demanded. "What are you trying to do, destroy my business??" I'm sure that today, when people hear tales of high school superintendents who insist that "If English was good enough for Jesus Christ, it's good enough for our students!" they know that they're hearing a joke.

However, the U.S. government and some state governments are now working seriously to establish "English Only" laws that would drastically restrict the use of foreign languages in the United States. Isn't it a bit strange to spend tax money for classes to teach Spanish to English-speaking students in our schools, while at the same time making it difficult (or illegal) for Spanish-speaking individuals to use Spanish there? This is just one of many curiosities resulting from the current confusion about language and languages.

Some accounts give the impression that general agreement now exists on multilingualism and that language policy decisions about it in the United States are made on the basis of firm knowledge and solid science. I wish that were true; it's not. Our feelings about bilingualism and multilingualism are a dense thicket of muddles and contradictions. As the six incidents on pp. ix-xi show, there is no one attitude that represents a national consensus, or even a national majority opinion. Our decisions with regard to language issues such as multilingualism, most of the time, are based not on scientific facts but on folklore and politics and emotion and expediency. We can't allow this to go on. In a world that is so rapidly becoming a global society, we have to do better.

My intention in this book is to take some useful steps toward sorting out this linguistic turmoil and laying a foundation upon which others can build to carry the improvement forward. Because there are so many differences of opinion, and because those differences are as likely to come from scholars and experts as from the general population, I won't be able to provide all the missing answers that would be needed for a rational and practical language policy. However, I will do my best to … 

• Set out the questions that must be answered with regard to the power of language in human life, especially in the areas of public policy and in education.
• Establish the importance of multilingualism within the context of the larger topic of the power of language.
• Explain why there is so much confusion and contradiction about multilingualism and about language learning as a whole.
• Provide a fair hearing for both (or all) sides of the major controversies.
• Present the basic facts, when they are available.
• Discuss the possible effects of multilingualism in some major areas of human life.
• Present useful information gathered from speakers of two or more languages about their personal perceptions of, and experiences with, multilingualism.
• Help make possible the building of a consensus for or against multilingualism that is based on scientific information and real-world facts.
• Outline possible directions for further research.

We will be focusing on seven questions—quite broad questions. They were chosen because I believe that the search for their answers will help us find ways of thinking about multilingualism and human language abilities that are more appropriate for today's world. They are:

1. What is multilingualism?
2. What is the link between languages and cultures?
3. What effect, if any, do languages have on human perceptions of reality?
4. What are some consequences of multilingualism in a few representative areas of human life (for example, medicine, business, and family life)?
5. Should we have a single language that everybody learns, worldwide, with which to carry on international communication? What effects might that have on our lives?
6. How do people become multilingual, and what happens in their minds and/or brains when they do? Is this good or bad?
7. What does all this information tell us about the power of human language?

A distinction is usually made between "bilingualism" and "multilingualism" when such questions are discussed. The reason for that distinction is not that we know the two "isms" differ in some significant and relevant way; we make the distinction only because it's traditional. It's also a nuisance. It's a nuisance to keep specifying whether we're talking about two languages or more than two. In this book the term "multilingual" will be assumed to include everyone with a command of more than one language, however many that may be, unless there's a good reason to give the exact number.

We will also be concentrating on spoken language, for the most part, although the many different situations among my respondents with regard to literacy in their various languages deserve a book of their own. (For example, one of my Latvian respondents—who has asked to remain anonymous—writes that the only way she can understand written Latvian is to read it out loud. When she hears it, she understands what she has said as she was reading it, but she can't read it just by looking at it.) Reading and writing are important in multilingualism, and will be brought up where directly relevant, but they're not our primary concern.

While writing this book, I have done the things any researcher and writer does in the course of writing nonfiction. I've read stacks of books and academic journals and newspaper stories; I've spent endless hours reading on the Internet. I've consulted with scholars and experts in the academic and scientific fields concerned with language learning, multilingualism, and related topics, and have profited greatly by their counsel.

In addition to doing that necessary groundwork, I have collected lengthy questionnaires from multilinguals all over the world who have graciously and generously given of their time and energy to help me with my task. Their questionnaires form a valuable database that is a window on the world of language as multilinguals perceive it. Each of my respondents speaks English plus at least one other language; sixty-nine speak English plus two or more. Altogether they represent fifty-five different languages—or dialects of Terran—from Afrikaans and Arabic to Xhosa and Zulu. (You'll find the complete list of languages on page [page numbers omitted].) Certainly it's important to know what scientists and scholars and medical experts think about the questions we are investigating here; it's equally important to listen to the ideas and opinions of people who have intimate personal knowledge about those matters.

Full Disclosure

Multilingualism is an issue about which such intense passions exist that I am reminded more than anything else of the uproar over the theory of evolution. People don't just have opinions about it, they hold those opinions devoutly. We haven't yet seen headlines about factions demanding the teaching and maintenance of multilingualism on religious grounds, in spite of the biblical story of the Tower of Babel—but I can easily imagine that happening. We've seen plenty of headlines about the scientific controversies and the political ones; those stories have been sensational, even inflammatory, and, far too often, misleading. Mululingualism is what's known as "a hot topic."

In the interests of fairness and full disclosure, therefore, I have to tell you where my own convictions lie with regard to the major controversies we'll be examining. I want to set out my own positions for you here so that I don't have to try your patience by coming back to them over and over again later in the book. And I want to acknowledge, in advance, that many important and distinguished scholars and scientists disagree with me. Here is my personal list of carefully considered conclusions, for your reference:

• I am convinced that Chomsky is correct in his claim that human infants are born with some innate capacity for learning human languages.
• I am convinced that there is a core of elements common to all human languages, usually referred to as universal grammar but more accurately global grammar.
• I am convinced that the link between language and culture is so strong that loss of a language inevitably means loss of the associated culture as well, and I consider such losses tragic.
• I am convinced that human languages (and dialects of human languages) do structure and constrain human perceptions of reality in significant and interesting ways.
• Finally, I am convinced that multilingualism is desirable for humankind and should be encouraged in every way possible.

I am indebted to so many people for this book that it's impossible to list them all. I would have to begin by listing the thousands of scholars who have built the foundation of knowledge on which I have had the privilege to stand as I wrote. I would have to list every teacher who encouraged my love for languages (and later, my love for linguistic science), and every speaker of a foreign language who was patient with my attempts to speak one. The lists would be endless, the task impossible. I will therefore mention only a few of those to whom I am deeply grateful, with the understanding that I'm well aware I'm only scratching the surface. My respectful thanks, then:

To Leonard Newmark, who showed me how to teach not only languages but a love for learning. To Amanda Cook, my editor, who has been a constant source of support and information. To my literary agent, Jeff McCartney, whose comments and insights as I worked through the many drafts and revisions have been invaluable to me. To my daughter and colleague, Rebecca Haden Chomphosy, who has been my research and fieldwork associate. To my multilingualism questionnaire respondents, without whom this would have been only half a book. To my colleagues in linguistics who have put up with my endless questions and shared their knowledge unstintingly when I needed it, even when they profoundly disagreed with me. And to my long-suffering family, who as always has had to endure the burdens that are inherent in living with someone who is With Book.

All of these persons—and the many others I have not identified here—are entitled to much credit and much gratitude; none of them is responsible in any way for my errors or omissions.

Despite bilingual nannies, tapes, picture books, and even sending their kids abroad for the summer, bilingual parents in this country are losing the battle to maintain their children's grasp of a foreign language … The challenge to raise a child bilingually becomes even more difficult for languages of lesser prevalence.

(Mireya Navarro, quoted in Landers 1996, page 5)

The quotation above describes a reality very unlike the one we ordinarily see described and discussed in the mainstream media. How could it be a challenge to raise a child bilingually or multilingually, as Navarro claims? All the news stories and editorials are about the dangers of multilingualism in the United States and the need for defensive strategies against it, as if American English might vanish at any moment! With the single exception of documentaries about Native Americans, we see no news reports in which parents are lamenting the fact that—despite their best efforts—their children speak only English.

We need to know which version of reality is accurate, and we need to find out how and why two such wildly differing portraits of our language environment can exist simultaneously. We can be reasonably sure that a communication breakdown like this one occurs only when two or more groups of people are using the same words and phrases but aren't giving them the same meanings. These situations which I call "reality gaps," can go on for many years without anyone who is involved even being aware that they exist; meanwhile, the problems they create are blamed on all sorts of irrelevant factors.

Our first and most basic question is: What do we mean when we say that someone is multilingual? And our first and most typical reaction to that question is of course that it's absurdly easy to answer. After all, aren't we're just talking about people who know more than one language? Isn't the meaning obvious?

The question isn't easily answered, and the meaning is anything but obvious. This is because we can't define "multilingual" or "multilingualism" until we have agreed on the meaning of that phrase, "know more than one language." No real consensus exists, you see, on what it means to "know" a language, or on precisely what a "language" is. We are well and truly stuck at this point, and the lack of agreement has created a seething broil of controversies. Controversies that can lead to the spending or withholding of many millions of dollars; controversies whose consequences can be as trivial as a bumper sticker—and as grave, even tragic, as civil disorder, terrorism, and civil war.

Defining "Multilingualism"—What It Means to "Know" More than One Language

David Crystal (1992, page 259) defines multilingualism as the ability to "make use of" more than one language. By this definition I am multilingual; I can "make use of" quite a number of languages. My Webster's has no entry for multilingualism, but defines bilingualism as someone's ability to use "two languages with equal or nearly equal facility." By this definition I fail to qualify, because although I speak French very comfortably, I certainly don't use both English and French (or English and any other language) with "equal or nearly equal facility." In Mirror of Language: The Debate on Bilingualism (1986), Kenji Hakuta tells us on page 4 that "very few people who would be considered bilingual have anything resembling native-like control of both languages." The lack of native control he refers to will only increase as the number of languages increases. Francois Grosjean writes that it is "perfectly normal to find bilinguals who can only read and write one of their languages, who have reduced speaking fluency in a language they only use with a limited number of people, or who can only speak about a particular subject in one of their languages" (Grosjean 1994, page 1656). It's common for multi-linguals in my database to judge their own language proficiencies in ways that reflect this confusion; for example, Karen Arouca writes that her oral fluency in Spanish is "not very good," but notes that when traveling in Argentina and Uruguay she is "sometimes taken for a native." How are we to make sense of all this?

The simplest way I know to clarify matters is to compare the language skills of my first husband (Peter Haden) with my own. His mother was French and his father was an American diplomat. Peter grew up in France and Switzerland in the international diplomatic community, a language environment where multilinguals were as ordinary as pondwater. It is an ordinary event in Switzerland for conversations to go on among three or four people, each of whom is speaking a different language, without anyone seeming to notice; such conversations are common on Swiss television. Peter learned both French and English from birth and had native fluency in both; the only difference he was ever able to identify for me was that he found it a little bit easier to do math in French.

As for me, I began studying French in high school and continued at the University of Chicago (where I met Peter), and I always got As. After we married, we spent some months living with his family in Geneva and in the south of France, speaking almost exclusively French during all that time. I have a bachelor's degree in French, and a high school teaching credential for it; I've taught French at both the high school and the university level. I worked as a French translator for a time in Washington, D.C. And Peter and I spoke French in our home after we returned to the United States. Does this mean that the one word "multilingual" can be used to describe both of us?

Well, it depends on how we define our terms, bringing us right back to where we began. Peter spoke both English and French as flawlessly as anyone speaks a native language. I speak English as a native, but my French is barbaric. I make grammar errors constantly. I use constructions that no native French speaker would ever choose, because they let me avoid some of the parts of French that I find difficult. My accent is an indescribably odd mixture of the French of Toulouse, the French of Geneva, and the French taught in Missouri's small-town schools in the 1950s. Nevertheless, if I were suddenly teleported into a community where French was the only language spoken, I could manage perfectly well. I am at ease carrying on conversations and handling daily affairs in French; I could teach in French without feeling uncomfortable. Perhaps most important of all, when I don't know how to say something in French, or don't understand something, I know what to say to repair the deficiency on the spot.

Peter and I are perfect examples of the two extremes of multilingualism. The linguistic difference between us should serve to make a point that is crucial to this book and must be understood: Statistics about exactly how many multilinguals there are in a given location are almost always meaningless. When Cambridge University Press announced the launching of a new journal (Bilingualism: Language and Cognition) on the Internet on March 24, 1997, its claim was that "more than half of the world's population use two or more languages in their everyday lives, and may therefore be defined as bilingual" (multilingual, in our terms). I welcome the journal, but I would have to ask: "Bilingual in what way? The way my husband was bilingual, or the way I am, or something in between?"

At the University of California-San Diego, where I taught in the French section of the foreign language program, it wasn't grades that got students through the classes. A perfect accent and flawless grammar wouldn't do it, nor would top scores on written exams. To satisfy the foreign language requirement for graduation, our students had to survive an exit interview—a ten- to fifteen-minute conversation—with two native speakers of the foreign language. The two interviewers had to agree that the student "knew" the language and knew it well enough to participate in that conversation with ease. Otherwise, the student—no matter what his or her grades were—did not pass the course and could not graduate. (Yes, many students did find this requirement almost impossible to believe, since grades are the measure by which most educational institutions decide whether people "know" a language.) By the UCSD standard, even if my interviewers had spent the fifteen minutes hilarious with laughter at my French, I would meet the definition of multilingualism.

Respondent Fran Stallings (English/Japanese) had an experience that casts a good deal of light on the difference between knowing a language in the sense of being able to use it in the real world and knowing a language in the sense of being able to pass tests. She was touring the United States with a Japanese scholar, and was asked (by the Japanese teachers at a training institute) whether she spoke Japanese. She demonstrated that she was fluent in the real-world sense of that word by answering with the standard phrase a native Japanese speaker would have used, a phrase that means "Just a few words." A teacher on the American team immediately demonstrated his own fixation on the "test-passing" sense of the word by a rude challenge: How, he wanted to know, could she possibly serve as the Japanese scholar's tour manager and translator, if her Japanese was no better than that? Dr. Stallings knew Japanese: her challenger only knew the answers to test questions.

Defining "Language"—What It Means to Know "a Language"

There is also no exact definition of "a language." The number of languages existing on Earth today is either roughly 6,000 or roughly 10,000 or roughly a figure somewhere in between those two. Depending. Barbara Grimes' Languages of the World (twelfth edition, 1992), a respected scholarly source, sets the number at 6,528. David Dalby, who directs the Observatoire Linguistique (a team working on a directory of the world's languages for the United Nations), announced in the July 29, 1997, London Guardian that the number of Earth languages is more than 10,000. This is a very large difference indeed, even for scholars. How are we to explain it? The cliché among linguists (usually credited to Max Weinreich) is that a language is a dialect with an army and a navy; that may be overstated, but it's true that the line is frequently drawn for political rather than linguistic reasons.

According to Jared Diamond, all of Europe has "only about 63 native languages," while New Guinea is at the opposite extreme with "about 1000 languages—yes, 1,000 mutually unintelligible languages" (Diamond 1997, page 544). The reason the word "about" appears in both those statistics is that other people might draw the lines between the languages and dialects differently.

I've worked as a linguist with both Navajo and Apache, and I would consider them dialects of a single language; I would consider a person who spoke both Navajo and Apache multidialectal, not multilingual. But neither Navajos nor Apaches, in my experience, would accept that judgment. Some linguists consider the variety of Laotian spoken by my son-in-law to be a dialect of the Thai language; but Bounlieng, who speaks both Lao and Thai, finds that idea unacceptable. No one (except perhaps the hypothetical Martian linguist mentioned in the preface) disagrees with the statement that Chinese and English are two different languages, but for many thousands of other "manners of speaking" on this earth it's impossible to draw the line that clearly.

Does it matter? Most people believe that it does. It makes us uneasy to think that we can't know with reasonable accuracy how many languages our planet shelters. Few of us view it with the concern felt by those peoples of the northwest Amazon who consider marriage between native speakers of the same language to be incest, but it matters a great deal nonetheless. I will be discussing ways we can be certain that it matters (as well as many ways that it might matter) as we go along in this book.

At this point it will be clear that there's only one sensible way for us to proceed. Our definitions of multilingualism are far too vague. Nevertheless, even though we have no solid definition for what it is to "know" more than one language, whatever "a language" may turn out to be, we have to keep moving. There's no other way to broaden our knowledge base and make progress.

How Many Different Languages Could One Person "Know?"

Suppose that we use our terms loosely (as they are most commonly used) so that by "knowing a language" we mean something like "being able to carry on conversations with other speakers of the language easily and comfortably." Put that way, how many languages might a single individual know?

Theoretically, as long as the languages are learned in infancy and early childhood, no upper limit exists. In practical terms, however, even if a child lived in a giant commune with native speakers of fifty different languages, it probably wouldn't be possible for the child to learn all fifty. There aren't hours enough in the day for adequate exposure to that many languages. And in practical terms, most human households even in societies where multilingualism is routine are composed of native speakers of only a few languages. Still, if someone learns "a few" as a child and goes on adding more as an adult, where is the upper limit?

We don't know. A recent discussion of this question on the Internet list for linguists had anecdotes about persons alleged to know sixty or more languages, but without any way to verify those numbers or know just what they might mean. A story by William H. Honan in the New York Times for December 31,1997, titled "To Masters of Language, A Long Overdue Toast," began as follows:

Each vocation or avocation has its Babe Ruths and Baryshnikovs, and so it is with the Linguistic Society of America. The champion is Francis E. Sommer, fluent in 94 languages … There were no Olympic medals or Nobel prizes for him, but that is a mistake that should be rectified, said scholars coming to New York City … for the annual convention of the Linguistic Society of America.

But suppose we accept on faith the claim that Francis E. Sommer was fluent in ninety-four languages. We have no way to be certain what that would mean. As Dorian (1982, page 33) says:

At the lower levels of semispeaker proficiency there are some individuals who actually say very little yet continue to interact in a highly successful fashion with fluent speakers. Thanks to their superb passive bilingualism they follow all verbal interactions without any difficulty, and … they can participate for years in a linguistic network without anyone paying much attention to the assymetric nature of their interactions, or even to their flawed productions.

I see nothing unreasonable, based either on the scholarly literature or my own experience, about the idea that people might be comfortable with scores of languages in the fashion that Dorian calls "passive" multilingualism—that is, able to function and carry on their daily lives in an area where any one of those scores of languages is spoken. If you and I are having a conversation and it seems to us that we understand each other, if neither of us is stopping every few minutes to ask what a word means or complain of being unable to keep up, it's very unlikely that either of us will issue some challenge that would require proof of fluency in the language. That happens only in special situations where professional services are being rendered, as when a teacher says to a student, or a lawyer says to a client,"Now, tell me what I just said, please, so that I can be sure you understood." Our foreign language examiners at UCSD all agreed that one of the most important demonstrations of "knowing a language" was that a student laughed in all the right places—a skill that has little to do with accent or verb endings.

In many parts of the United States people tend to look upon multilingualism as something strange and exotic. Unless they work in a field where multiple languages are a standard professional requirement, they may not know personally even one person who speaks more than one language. They are likely to assume that because this is true of their personal language environment it represents what is usual and normal everywhere. Their assumption would be incorrect; in much of the world, it's monolingualism that's unusual, not multilingualism.

James Geary writes that it's common for many Africans "to speak five or six tribal languages in addition to their own mother tongue" (Geary 1997, page 6). One of my husband's relatives was fluent in seven languages, as a result of having spent his childhood in a missionary compound where all seven were spoken. In the foreign language program at UCSD it wasn't unusual for people to be very fluent in four or five languages; one of my associates there was fluent in fourteen.

In my database of multilinguals, people who claim to be able to get by comfortably in four or more languages are common—and many take their expertise with English so much for granted that they didn't even list English as one of their languages when they filled out my questionnaire in English. If I were to create a sort of "Prototype Multilingual" from that database, the linguistic life story would go something like this:

My father spoke Polish and my mother was Italian, and we spoke both of those languages at home because my grandmother lived with us and her only language was Italian. Then when I was three years old we moved to France, and it took me about six months to learn French, which was the language all my friends spoke. I started taking English in school when I was ten, and then we moved again, this time to Turkey, and so I learned Turkish. My first job was with a company that sent me to lapan, and so I learned lapanese when I was in my early twenties. Somewhere along the way I picked up German—I don't really remember when—and although my Russian certainly isn't native I can get by well enough for social and business conversation in that language as well.

How Hard Is It to Become Multilingual?

We have more than enough evidence that it's no real stretch for tiny kids to learn half a dozen languages natively if they're regularly exposed to all of them; the number that they can learn easily may be much higher than that. Matters are different for adults, whose language learning performance tends to vary widely, but we can be reasonably certain of a few basic items:

• The hardest foreign language for an adult to learn will be the first one learned, all else being equal.
• Each tongue learned thereafter from the same language family will be easier to learn than the first foreign language was.
• Even languages from different language families—for example, Spanish and Japanese—will be easier to learn after a few other languages have been mastered.
• So far as we know, and so far as time allows, any number of languages can be learned over a lifetime; no evidence exists to support any upper limit.
• With rare exceptions, none of the languages learned as an adult will be learned with native fluency.

When linguists are faced with the need to learn a new language, they're not alarmed, even when it's a language for which no written form exists. It's not just that practice makes perfect. It's that the characteristics of human languages, in spite of their unarguable differences, are a relatively limited set. Writing of Noam Chomsky, David Berreby (1994, page 49) tells us that:

he likens language to an elaborately wired box of switches—root concepts and grammatical principles—set in one position or another by experience. Chinese sets the switches in one pattern, English in another. But the fundamental conceptual framework of how concepts and syntax interlock is the same.

Theoretically, any sound a human being can make could be part of a human language. Why not? In actual practice, however, all languages choose their sounds from the same set of possibilities, a set containing somewhere around seventy to eighty sounds. Nothing logically—or physiologically—prevents a language from having words that begin with a sound made by touching the upper lip to the lower teeth. Many languages have sounds made with the upper teeth touching the lower lip, like English "F" and "V." But no language uses that other option. A language could, logically, include as one of its sounds the sound of the hands clapping, or the nose sniffing; no languages take advantage of those possibilities. Two languages may be separated in time and space by vast distances; nevertheless, they will confine their choices of meaningful sounds to the basic human set.

In the same way, nothing prevents a language theoretically from having any grammar rule the human mind can imagine. However, all human languages use only four kinds of grammar rules:

• Rules that delete something. (We can say either "You eat your peas!" or "Eat your peas!")
• Rules that insert something. (We say, "It's raining," although there is no "it" that rains.)
• Rules that substitute one thing for another. (Instead of "I was looking for the book about Indo-European languages and I found the book about Indo-European languages," we ordinarily say, "I was looking for the book about Indo-European languages and I found it.")
• Rules that move things from one place to another. (We can say either "I hate eggplant" or "Eggplant, I hate")

You can reduce that to three kinds of rules if you define deletion as the substitution of zero.

Within that set, we can think of an infinite number of things—sounds, words, parts of sentences, and more—to delete or insert or substitute or move, and we're fully capable of using all sorts of curious rules about those processes. Just consider the many "secret languages" that human children all over the world are so fond of, like English pig Latin.

There's no logical reason why a language couldn't make a sentence negative by repeating every word it contained. That would be an insertion rule, easy to understand and easy to use: "After each word, insert a copy of that word." If English did that, the negative of "The kids ate pizza" would not be "The kids didn't eat pizza," but "The the kids kids ate ate pizza pizza." No problem! But no language uses that option, and no language has a rule that repeats every word in the sentence, not for any purpose.

Because the characteristics of human languages are limited in this way, the more languages you learn, the more items from the set of possible characteristics you will already have come across in some other language. French forms its negative sentences using two separate chunks, "ne" and (among other things) "pas." When I studied Navajo, which is about as unrelated to French as a language could be and still be Terran, it was easy for me to get the hang of the Navajo two-part negative formed with "doo" and "da." I was already familiar with that way of doing things from French, and it came as no surprise. As you add languages, you think, "Oh yeah! They do that in Kwakiutl!" and "Sure, I get it—it works like Korean." When you approach a language that has never been written down before, as linguists are trained to do, you already know many many things about it in advance, simply because they are things that are true of every human language.

When science fiction writers try to write about extraterrestrial (hereafter, ET) languages, they can throw letters around and create an impression of something truly exotic and different. Some rely on clots of letters and have their ET speakers say things like "Ff'tt'qlwx q'wxwxllp!" Some of the science fiction tongues are amazingly ingenious. But when these languages are carefully examined they always turn out to contain only characteristics that can be found in human languages. This isn't because the writers lack skill or imagination. It's because the definition of what counts as a language is part of being human.

How much of language is innate—"hardwired" in the human brain—is a matter of passionate argument; I've only touched on it here. (We'll come back to that argument in Chapter 8.) But there is wide general agreement that the specifications for what can qualify as a human language are innate. Presumably you would have to have a nonhuman brain to be able to imagine and describe the characteristics of a truly ET language.

Language Learning in Young Children Versus Language Learning in Adults

Knowing these things should make it easier to understand how tiny kids can learn two or more languages at the same time, a task that most adults would find intimidating. Children are born with the full set of language possibilities either hardwired or available to them in some other fashion. As they're exposed to language in their environment, they begin eliminating the possibilities that aren't used in the language they're observing.

Consider what this means in terms of the way languages sound, since sound is so much of what's involved in having a "foreign accent." When babies start babbling they make sounds that occur in some human languages but that don't occur in the speech of the people in their language environment. Babies in an American English household who babble the Spanish "B" sound or the French "R" sound but never hear either one used around them quickly give up making those sounds; if they try to learn Spanish or French as adults, they usually find those two sounds hard to pronounce correctly. Babies in whose surroundings both French and Spanish are spoken, however, will never drop that "B" and "R" from their inventory of sounds that they recognize as worth hanging on to. The closer a child is to babyhood, the more recently foreign sounds will have been discarded; it's not surprising, therefore, that they're easier for young children to learn than for adults.

We don't know what effect it has on tiny kids to be exposed to a language and then have that exposure end very early in life. When my first husband died, my youngest child was only three years old. Her exposure to French in our home ended at that point because I had no desire to provide my children with regular exposure to my own peculiar French when no more standard model would be available to them for comparison. When Patricia studied French in college she had to learn it from scratch, but with one somewhat unusual characteristic: French sounds that are almost without exception difficult for English speakers were easy for her, and her accent was excellent from the very beginning. Perhaps this was because hearing French at home during the first three years of her life did have an effect, even though she had no conscious memory of the language; perhaps it was only a coincidence. No evidence is available for choosing between the two possibilities.

At one time we took it for granted that although infants and toddlers could acquire more than one language with ease, it would cause an inevitable delay in their "language development." They would start talking later, we believed. They would mix up their languages. They would put together sentences like the ones that respondent Rhonda Eikamp reports for her three-year-old daughter growing up in the family's English/German household in Germany:

"Ich habe das aufgefmisht." ("I've finished it up.")

"Er supposed das nicht machen." ("He's not supposed to do that.")

The more languages there were involved, the more severe the delays and confusions were expected to be. All this was something "everybody knew." Given the problems most adults in the United States associate with foreign language learning, it all seemed to be self-evident.

Even if the claims had been proved true, I'm not sure they would be anything to worry about. However, dissenting reports have steadily piled up from parents of multilingual children, saying that their kids prove the predictions of language troubles false. It appears that children who have serious or long-lasting problems are the exception rather than the rule. Hakuta tells us: "In most children by the age of three to four, languages are rarely mixed" (Hakuta 1986, page 67). And parents agree: It's unusual for any problems caused by confusion among the languages to last much beyond that age.

One of the most striking things about the information in my database of multilinguals is their relaxed attitude toward their multilingualism. Not a single one complains of having been held back in language development in any way. None report any anxiety about language development in their own children (or look upon mistakes like the ones in the example above as anything serious). And they are unanimous in wanting their own children—existing or potential—to be raised as multilinguals. Typical comments:

"I want them to learn as many languages as they can." (Bertil Eroz, Turkish/English)

"I think children should be exposed to as many languages as possible and given the opportunity to hear them when they are still young enough." (Dominique Estival, French/English)

The closest thing to a negative comment about bringing kids up multilingually comes from Marie Lenstrup (Danish/English/ Swedish/Norwegian, living in Hong Kong), who says, "I'm not sure that it's such a great idea for little children to have too many languages to choose from." But she goes on to demonstrate that what she means by "too many languages" is four or more, saying, "I would of course want any children to speak both my language, the language of their father, and the language of the place we were living in." All my respondents appear to take it for granted that their children should learn at least two or three languages, just as the typical U.S. parent takes it for granted that they should learn to read and write.

Contrast this with the statement from Ruey Brodine (English/Italian): "The process of learning a new language and a new culture are so difficult that I can't imagine anyone undertaking it without having to. But if you stick to it, no other learning experience is more enriching, in my opinion." Brodine (who lives and works in Italy) didn't begin learning the Italian language and culture until the age of twenty-six, by which time it often is a formidable project to learn a first foreign language well. Small children ordinarily react to language learning in this way only when they've been overexposed to older persons' complaints that the process is something to dread.

The idea that learning a language will always be an ordeal is very common among native American English speakers in the United States, even for people who strongly approve of it as a project. Walter Meyer accurately points out in his wonderful book Aliens & Linguists (1980, page 117) that American science fiction writers will cheerfully put their characters through almost any imaginable sort of awful ordeal, and some that almost aren't imaginable, with one exception: Characters are spared the horror of having to learn a foreign language. The Star Trek series reflects this attitude by seeing to it that all potential language problems are settled with a flashlightlike (and extremely improbable) gizmo called a "Universal Translator." If I point my UT at you and you speak Southern Colloquial Venusian Tqiltqish at me, I will hear American English; and when I answer you in American English, you will hear it as Southern Colloquial Venusian Tqiltqish. This would save huge amounts of money at the United Nations and international gatherings of all kinds, and is without question necessary to keep the filmed stories moving along, but it's fantasy rather than science.

Current Attitudes Toward Multilingualism

Three attitudes toward multilingualism prevail in the United States today. It's not unusual for someone to hold all three of them at once, seemingly without any awareness of what remarkable mental gymnastics that requires. They are:

1. Knowing more than one language is a mark of accomplishment, something to be admired.
2. Knowing more than one language is a transitional condition, a burden that has to be put up with until English is learned and the other language(s) can be abandoned.
3. Knowing more than one language is a mark of some sort of deficiency, whether intellectual, educational, social, or other.

Multilingualism Is an Advantage and Should Be Admired

The alleged benefits of knowing more than one language, from the point of view of the American English speaker, are well known. Multilingualism broadens the mind; it improves your ability to learn all academic subjects, even those that have little to do with language. It makes you a more well-rounded person; it gives you more than one "worldview." It equips you for success in the same way that any other array of skills does, like the circumstances of the multilingual lawyer on page x whose colleague is afraid of losing him because with fluency in four languages "he can write his own ticket." It is in every way a good thing and will improve your life, goes this school of thinking. So much so that it was once a mark of real wealth: Wealthy families would hire live-in French governesses and Italian tutors for their children so that they could acquire those languages as part of daily life, and they would see to it that their youngsters (at least the boys) were taught Latin and Classical Greek in childhood as a matter of course.

I often talked to my university students (in the 1970s) about this rosy picture of multilingualism. They usually agreed with it; many said they wished they'd had that kind of childhood education themselves. But when I would ask them whether it made a difference which foreign languages children learn, they were quick to give me their personal list of languages that don't broaden the mind. "Why not?" I would ask them. And they would look at me with surprise and say, "Well, everybody knows that!"

Multilingualism Is a Burden—Something to Get Rid of as Quickly as Possible

The alleged burdens are equally familiar. Multilingualism will hold you back in your language development and make it hard for you to succeed in school. It will confuse you and make it impossible for you to fit in with your peers. It will keep you from assimilating to the culture; it will "ghettoize" you and "keep you from becoming an American." You're likely to end up on welfare, or worse. It is in every way a bad thing and will have negative consequences; it may ruin your life. My students were dubious about the evidence for this, but not because they had any evidence against it (although some suggested "the mess that it's made in Canada" as a reason for their feelings). For most, it was once again something that they "just knew."

And how do my multilingual respondents, who presumably would have had to struggle with these burdens, feel about it?

My experience with knowing another language is that the world is much bigger, and more interestingly multifaceted. I know a few multilingual people, and they seem so … large (Elisabeth Vonarburg, English/French—both European French and the French of Quebec).

It's like being able to step outside yourself and take a look, or step outside the world and take a look (Rhonda Eikamp).

The more languages a person knows, the more access one has to alternate realities and the more imagination can rove. Multilingualism is always an intellectual advantage (M. J. Hardman, English/Spanish).

None of my respondents complained about multilingualism having held them back, having interfered with their success, or anything remotely of that kind. Where there were complaints, the typical statement was "I only wish I had learned a few more languages." Rebecca Haden Chomphosy has been talking to Thai and Laotian parents here in Arkansas, and they have without exception expressed the wish that their children should grow up speaking Thai (or Lao) in addition to English. None have said that they hoped their children would learn only English because it would be undesirable for them to know both languages.

In the interests of fairness (and science), I have to point out the possibility that when I sent out my call for volunteers to complete the questionnaire, the only people who responded were people who had never found multilingualism a drawback. Maybe there were five times as many who would have taken the opposite position, but none of them offered to participate. (It's not the case that only people with computer skills filled out questionnaires, by the way. Quite a few were passed along to people who filled them out with pen and pencil.) As is typical with preliminary data-gathering research of this kind, there's no way to know whether my group had succeeded because of their multilingualism or in spite of it, or if the two things are totally unrelated; much more research of this kind needs to be done.

It's also important to mention that after M. J. Hardman wrote that multilingualism is "always an intellectual advantage," she added, "This advantage can be, of course, and frequently is, nullified by negative social circumstances (like putting all Spanish speakers in the speech defect class, like they used to do in New Mexico when I lived there in the 50s)."

Multilingualism Is a Deficiency

The third attitude—which as recently as the 1950s was not simply a popular idea but a common official policy in education and social services—is no longer "politically correct" in the United States. It will be worded more subtly when expressed in public today. Still, it continues to thrive in the general population. (Read Hakuta's The Mirror of Language for a thorough and detailed historical account of the phenomenon.) When attorney Wendell Gautier was growing up in southwest Louisiana, he and his sisters and brothers were forbidden to speak Cajun French—his parents' native language—at home, because his mother thought "French-speaking children faced disadvantages" (Button 1996, page 82). I've recently read two op-ed pieces in national magazines written by young parents saying that they deliberately use only English around their children because they feel that monolingual English speakers have an advantage. The politicians who are so belligerent with their claims that "My ancestors were immigrants, and nobody ever gave them any bilingual education classes!" are presumably unaware that many of those same ancestors were labeled mentally deficient at school because they didn't speak English.

However, although the attitude has been toned down a little and is discredited among scholars, it has by no means disappeared from our educational system. Children with nonnative English skills (or whose dialect of English isn't one of the fashionable "Standard" ones) are still being "tracked" into vocational programs instead of being prepared for college. They still are being almost automatically excluded from "gifted and talented" courses and programs. Often the judgments depend on which foreign accent and grammar the child has. French (European French, not Cajun French) is fashionable and a French accent is "charming," but Spanish enjoys no such advantage. The fact that the people making such judgments often have no conscious awareness of what they're basing them on doesn't make the effects on the children any less negative.

Hakuta claims that "a bilingual person must have two parallel systems, one for each language, that must at the same time be interrelated" (Hakuta 1986, page 3). This is, he says, a "packaging puzzle" in which "two language-bounded mental and social systems must be housed in a single mind" (ibid.). And if there are three (or more) languages, we're talking about three (or more) "language-bounded mental and social systems housed in a single mind" (ibid.). It's easy to write that down and read it, but what does it mean? Is it a literal statement or a figure of speech? Does it matter?

Most important of all, should we care? In today's world of huge intractable problems—crime, plague, famine, war—should we bother to explore this question? I strongly believe that we should; I believe it would have critical implications for all of those intractable problems. In today's world, which is fast becoming a global society—perhaps turning from a motley collection of nations into a world called Terra—the most urgent question may well be whether we can afford not to care.

Some Necessary Definitions

To avoid confusion as we move on to the next topic for discussion, and to make that discussion less awkward, I want to close this chapter by establishing a few more definitions of terms.

Dialect and Language

When people perceive themselves not only as speaking the same language but as speaking it "the same way," they are using a dialect. A language is a group of dialects sufficiently alike that all who use them perceive themselves to be speaking the same language, although not necessarily in the same way. The exact line between dialect and language, as we have already seen, is difficult to draw. As an example about which there is a general consensus, we would all agree that English is a language, and we would all agree that the typical English of Brooklyn, New York, and the typical English of Little Rock, Arkansas, are two different dialects of English.

Native Language

Usually, the language that someone learns first, in infancy and early childhood, is called the native language. In Wolof, the metaphor for a native language is mother's milk; when a Wolof native speaker is asked to identify his or her native tongue, the most typical answer means, literally, "I was nursed on Wolof" (Swigart 1992, page 14).

My database for this book includes a few multilinguals who learned some Language X in this way, but who consider Language Y their native tongue because they abandoned Language X in the first few years of their lives—when their family moved to another country, or they lost the caregiver who spoke Language X, or something of that kind. Mikala Joergensen, for example, writes, "For a period between one and six years I was bilingual in Danish and Greenlandic, but I forgot Greenlandic when my family returned to Denmark when I was six. I no longer speak it." But this is unusual. For most of us, the language we learned first will always be the one we identify as our native language.

The terms "first language," for the native tongue, and "second language" (often used to refer to a third, fourth, and so on) are commonly used in this context, but my respondents objected strongly to that practice. Suppose you are born into a household where your father speaks French and your mother speaks Japanese, with people who speak those two languages coming and going all the time. Suppose you have an Italian nanny who looks after you while your parents are at work. Suppose also that the household is located in Hong Kong, so that as soon as you begin making friends outside your home you learn some variety of Chinese. In such situations, my respondents tell me, assigning labels like "first" and "second" to their languages is impossible. They refused to try. In this book; therefore, we'll confine ourselves to "native"—including circumstances in which more than one language is learned as a native tongue in infancy—and "other."

Pidgin and Creole

When two languages come in contact and their speakers must interact with one another, they sometimes develop a mixed language for that purpose that uses some features from each of the two tongues; such languages are called "pidgins." There may be more than two languages involved. Sheng, a language spoken by young people in Kenya, is a mixture of English, Swahili, German, and Latin—and Swahili itself began as a mixed language. Unfortunately, "pidgin," and often "creole" as well, are terms that have negative connotations for many people. It's important to understand that "pidgin" is a neutral term which is used to refer to "any language that is historically a mixture of two or more languages and that serves as a lingua franca for groups of speakers but is not the first language of any people." (O'Grady et al. 1989, page 344). When children learn a pidgin as their native language, it will begin to develop as any other language does. If it survives, it will then be called a "creole."

Variety

All the usual terms that refer to some "manner of speaking a language" have connotations of one kind or another, often unpleasant. My own native dialect is Ozark English. When people hear the phrase "Ozark English dialect" they think of the Beverly Hillbillies television show. They think of those tacky placemats you see in truckstops with the phony "translations" of Ozark speech for tourists. I'm very personally familiar, unpleasantly so, with the negative ideas that are associated with the word "dialect." There are times when we need a term that means, "I am referring to this way of speaking a particular language, and I am making no value judgments of any kind about it when I do so." "Variety" is that term.

Register

Some ways of speaking a language are distinct because they're associated with a particular role of the person who's speaking. The way that doctors typically talk while "doctoring" in the United States is an English register often called MDeitySpeak. The child who says, "Bye, Mom!" to his mother and "Goodbye, Mrs. Mason!" to the school principal and "Later, man!" to his friend on the playground is demonstrating command of three registers. The sequence below is a handsome example of Academic Regalian, a register commonly used by college professors:

Taken literally, the statement "I was wondering if you would be able to drive me to the airport" is a prolix string of incongruities. Why notify me of the contents of your ruminations? (Pinker 1994, page 230)

Sometimes languages themselves become very like registers. In Wales, for example, Welsh is the language of home and church and friendly social situations, while English is the language of government and business. In Maalula, Syria, the language used in schools and public life is Arabic; in the home and on the street, however, people speak Aramaic—the language that Jesus is believed to have spoken. In the United States, the same thing has happened with Mississippi Choctaw and English; Choctaw is the language reserved for private life and Choctaw-community-only events. The technical term for this situation, in which speakers of two languages assign each one to specific roles in life, is "diglossia."

Code Switching

When a multilingual person moves back and forth between two or more languages (or two or more varieties of a language) in a single communication event, that is called "code switching." For example:

With my fellow Hispanics … I can relax into our nuevo creole. (Fernandez 1986)

"Maa ngiy a genoux." ("I'm on my knees.") (Quoted in Swigart 1992, page 19)

Notice that in the first example we don't consider the word "creole" foreign; it has been borrowed into English and has become part of the English vocabulary. Nuevo, on the other hand, is the Spanish word for "new" and is definitely not part of English. In the second example, the speaker starts out in Wolof and ends with the French a genoux.

Both of these examples show code switching happening inside a single sentence. It also happens within longer sequences of language, with a sentence or two of one language, a sentence of another, back to the first language, and so on. In multilingual families a child may speak one language to a parent, another to a grandparent, and still another to a sibling; if all are involved in a single conversation, this code switching will go on the entire time. The languages don't have to be of the same language family, nor do their grammars have to be similar; for example, Japanese/English code switching is a common manner of speech in the Japanese community of Toronto.

Two things about code switching need to be kept in mind. First, it doesn't mean that the speaker is making mistakes or has the two languages or varieties mixed up (like the small child on page 14), and it isn't random. There are rules that govern code switching, just as there are rules for the use of a single language or variety. Spanish/English code switching allows "Es muy friendly," in which the Spanish for "very" is followed by an English adjective, but "Es very amable"—English "very" followed by a Spanish adjective—is forbidden. Second, in any multilingual/multicultural society, code switching is a valuable skill; the more codes someone knows, the more choices that person is likely to have in life.

Code switching isn't usually a term applied to the act of moving among various registers of a single language, but it should be. We've all watched doctors on television droning on in the medical register while an interviewer pleads with them to "speak English"; their inability to switch out of a register in which kidney stones are known as nephroliths is a very real handicap. Highly skilled attorneys switch registers at the drop of a hat, to make themselves seem either more or less threatening to those in the courtroom they're speaking to, depending on their strategic need of the moment. The ability to switch from one register to another at will, according to the language environment around you, is like being able to play any instrument in the symphony orchestra; it removes barriers in all directions.

Language Policy

The term "language policy" refers to the official positions that governments take with regard to language use and maintenance, whether supported by actions—laws passed, programs funded, grants made, and the like—or supported only by rhetoric, in speeches and position papers and public statements. At the moment, unfortunately, neither the U.S. government nor any of the governments of the fifty states (with the possible exception of Hawaii) can be described as having a coherent language policy.

The most basic reason why linguistic diversity should be preserved is that language helps people to retain their culture.

(Kleiner 1995, page 15)

What new immigrant families across the nation are learning, as their predecessors did before them, is that the power of American culture, and particularly the lure of television, is so strong that it is a challenge to raise a child who can speak a foreign language fluently.

(Navarro 1996, page 5)

A culture is the sum of all the things that make up the way a group of people live their lives: how they dress, work, eat, study, raise children, respond to illness, create art, worship, play, argue, make love, function within a family, perceive time and space, and so on. That all peoples have cultures and languages, both of which are of tremendous importance to them, is not a matter of dispute. But there is a debate about whether the language spoken within a culture is just one more element like clothing or housing or music—part of the culture but separable from it—or something quite different and not separable.

A common consequence of the idea that language and culture are inseparable is a conviction that any change in the language threatens the culture. The French have an official language academy to watch over linguistic developments and try to regulate them. They're so alarmed by what they consider to be the "contamination" of French by foreign tongues (especially American English) that they've resorted to drastic measures. French courts recently overthrew a law that would have "forced private-sector TV stations to eschew borrowed English words—le weekend, le leader, and so on—that are sprinkled through every newscast" (Andrews 1995, page 134); the courts ruled the law unconstitutional. That hasn't stopped the language guardians. Their current strategy is to set up for all the mass media obligatory percentages of content created and produced entirely in France—that is, to set up cultural quotas.

There's an Internet site called Maison de la Françité (roughly, "House of Frenchness," with a nice pun included) run by le Cercle de qualité du français dynamique (CQFD). Here you can find proposed "pure French" replacements for foreign words, with detailed explanations and justifications. For example, to avoid the potentially contaminating English "flipchart," CQFD proposes tableau-feuilles. The stated purpose of CQFD is to watch out for "Anglicisms" and similar language hazards "en vue de contribuer à la qualité de notre langue dans un monde en mutation rapide." (That is, to serve as a kind of quality control for French in a rapidly changing world.)

The French are far from being alone in such practices. Lithuania, for example, has "a force of paralinguists who prowl the streets for illicit phrase pushers. Perpetrators face warnings, fines and humiliation-by-radio" (Newman 1993). We are all familiar with the ongoing controversies in Quebec over laws forbidding English signs and the like.

The United States has no such laws—but not because no one has ever tried to establish them. In 1780, John Adams argued passionately (and unsuccessfully) for the creation of an American language academy to keep American English pure, and proposals for similar measures are a constant feature of editorials and letters to the editor to this day.

The Japanese are an interesting contrast. Despite being one of the world's most conservative cultures, they require their children to begin learning English in elementary school as a matter of course, and they borrow English words with great enthusiasm. They appear to have no fear of "contamination." Perhaps this is because their society is so uniform; perhaps it is because of the well-known Japanese conviction that everything Japanese is inherently superior; perhaps there are other reasons.

The truth can be bluntly stated: Every attempt to control spoken languages by laws has failed. If it can be done (and I don't believe that it can), we haven't yet figured out how to go about it. Controlling written language is to some extent possible, but all that happens when the written form isn't allowed to change is that the distance between it and the spoken forms of the language grows ever more vast over time.

The ultimate language change is of course its complete replacement by some other language. Our Founding Fathers considered this move shortly after the Declaration of Independence, "when the aversion to all things English extended to the language." They seriously considered making German, French, Greek, Hebrew, or even an artificial language the "official" language of this country. But a careful look at what would be involved led to a change of heart, summed up by Continental Congressman Roger Sherman with "It would be more convenient for us to keep the language as it is and make the English speak Greek" (Time, 9/18/96, page 38).

That's a funny story—but not every American will find it amusing. Many Native Americans in the United States are trying desperately to keep their languages alive, not only because they cherish their native tongues but also because they're convinced that if the languages die the cultures will die with them.

The state of Oklahoma, with thirty-five native languages and more than 76,000 Native American students, leads the country in this effort, especially for Cherokee, Choctaw, Kiowa, Creek, and Comanche. "With twenty-two Head Start centers introducing three-and four-year-olds to their native tongue, the Cherokee program has become a model for the entire nation" (Bilger 1994, page 41).

The Keresan speakers of the Cochiti Pueblo in New Mexico are equally concerned about their language, but have taken an unusual path. They are teaching their youngsters Keresan only orally, by the method of total immersion in the language, and will not allow it to be written down for learning. Mary Eunice Romero, who directs their language program, explains that "The language is sacred. One way to protect ourselves is in a sense to be exclusive" (Linthicum 1998). Whether Keresan will survive as an unwritten language in a world where literacy is taken for granted remains to be seen.

On the other hand, there are native peoples who are deeply offended by the suggestion that their cultures depend for their existence on the maintenance of their native languages; there are also peoples who are divided on this issue. Navajo Sylvia Wadsworth says, "Our Navajo language is who we are"; but Navajo Byron Charley says, "I think you would still be Navajo even if you're not a Navajo speaker … Being Navajo depends on how you conduct yourself and go about your life"(Cantoni 1996, page 133 and page 135).

Almost all the Navajo children entering kindergarten in the early 1970s spoke Navajo and few spoke any English. Today—despite the fact that the Navajo have created and maintained some of the most highly acclaimed native-language educational programs, including programs at the college level—things have changed drastically. About half of the kindergarteners speak only English, and the status that goes with speaking Navajo has been declining steadily among the younger generations. Whether there is any link between this change and the new epidemic level of violence among young Navajos is not known; I would be willing to wager a large sum on the existence of that link. (For a fascinating, and in many ways heartbreaking, account of the attempt to incorporate a Navajo teaching philosophy into the college curriculum, see "Diné College Struggles to Synthesize Navajo and Western Knowledge," by Paul Willeto, in the Fall 1997 issue of Tribal College Journal.)

Politicians in the United States point to examples of turmoil caused by multilingualism elsewhere in the world and demand that every immigrant begin learning English and carrying on daily life in English, if not from Day One then certainly from Day Two. An editorial in the October 22, 1977, Oakland Tribune described the situation like this:

To many, American English fluency is the rite of passage into being an American. To others, who generally agree that the ultimate goal is to be American English fluent, rigid time-tables for learning the language create more harm than good. Wrapped around all this, of course, is what some see as the real fight over preserving a single American culture or fracturing into the kind of multiculturalism that leads to ethnic Balkanization.

Some languages have been so loved or hated as vehicles for culture that they have become a justification for subversion and worse; think of Basque, Irish, and Afrikaans. At the same time, placid little multilingual Switzerland serves as our universal symbol of a peaceful and orderly nation; its languages and cultures appear to coexist and thrive without any serious discord. Clearly, there is no more an international consensus on this question than there is on the definition of multilingualism.

When you bring up the hypothesis that losing a language means losing a culture, people typically respond in four very different ways:

1. "Right! And we must not let that happen!"
2. "Right! And the sooner that happens, the better!"
3. "Nonsense. Language is one thing; culture is something entirely different!"
4. "So what?"

Let's consider these responses one at a time.

"Right! And We Must Not Let That Happen!"

When I speak English, I lose in a way part of my Spanish nature.

(Questionnaire respondent)

I teach language classes, and when I look at the students in there, I have to think about what we're giving them. We're trying to give them a language that will allow them to think in certain values, and certain lifeways that are important to our people.

(Parris Butler, teacher of the Mojave language, in Hinton 1995)

People who respond this way believe that language and culture are interdependent and inseparable. Many in this camp believe that every language and its culture, without exception, are precious and important and should be preserved. Linguist Ken Hale says that languages "embody the intellectual wealth of the people that speak them. Losing any one of them is like dropping a bomb on the Louvre" (in Bilger 1994, page 19). For some this response may well be based only on romanticism; when you get beyond that predictable reflex, the most common justification is based on the comparison of endangered languages with endangered species.

Like most metaphors, this one isn't perfect, but it comes very close. We can look at an endangered plant in a rain forest and insist that it must be protected because, for all we know, it's the cure for Down syndrome or the missing link in a scientific problem that will never be solved without it. In the same way, we can look at a language that has only a handful of living speakers, and insist that it must be protected—because, for all we know, it holds the key to ending domestic violence, or is the missing link in some crucial problem in linguistic science.

A scientific Foundation for Endangered Languages was founded in 1995. Its members named their newsletter latiku, after the mother goddess of the Acoma tribe of New Mexico, who "caused people to speak different languages so that it would not be so easy for them to quarrel" (Gill and Sullivan 1992, page 5). The international organization called Terralingua: Partnerships for Biolinguistic Diversity makes the language/biology metaphor explicit. Its mission statement includes, among other things: "To help preserve and perpetuate the world's linguistic diversity in all its variant forms" and "To illuminate the connections between cultural and biological diversity."

In 1997, the Endangered Language Fund (a nonprofit organization whose goal is to preserve languages threatened with extinction) awarded ten grants to projects that they judged likely to contribute to achieving that goal. Among the awards were:

• Funds to produce original television dramas in Choctaw and Creek, to be shown on cable access channels and made available as videotapes to native speakers of these languages in Oklahoma.
• Funds to help a linguist record the last two living speakers of the Klamath language, so that the recordings will be available to the Klamath tribe.
• Funds to buy equipment needed for language immersion programs in the Micmac, Maliseet, and Passamaquoddy languages.
• Funds to help prepare language materials such as textbooks and a dictionary for the Jingulu language of Australia, which now has no more than ten living fluent speakers.

Let's suppose for the moment that this first position is correct and that we therefore really should do everything possible to preserve every language and culture. What would that mean, in practical terms? To make the question more manageable in size and scope, let's consider only what it would mean for the United States.

American English (by which I mean SAME—Standard American Mainstream English) is not endangered. On the contrary; it's well on its way to taking over the linguistic world. We'll come back to that phenomenon in Chapter 7. But the native peoples who were here be fore the Europeans arrived had hundreds of different languages, and for them the situation is very grave.

Statistics on the exact number of native languages in the United States today vary (for the reasons set out in the Introduction, as well as because they are dying so quickly); according to linguist Leanne Hinton, there are more than 175. California alone has nearly 50 different Native American languages, and not even one is spoken by children or young people. In 1994, in Oklahoma, the last living speakers of Miami, Peoria, and Quapaw died. James Brooke reports that 55 of the Native American languages are now spoken by fewer than ten elderly persons, and 70 are spoken almost entirely by the elderly; he claims that only 20 languages remain that are spoken by all the generations, with children actually being raised in the language (Brooke 1998). Steven Pinker warns us in the gravest terms that "Languages are perpetuated by the children who learn them. When linguists see a language spoken only by adults, they know it is doomed" (Pinker 1994, page 259).

This information is critically important to our discussion, because the situation of these languages is a special one. Suppose laws forbidding the use of French in the United States were passed; suppose rigorous attempts were made to enforce them. That wouldn't be trivial. The Cajun French culture would suffer negative consequences, for example. But French itself would not be extinct. French is spoken all over the world, and children of French ancestry could go to other countries to experience it. This is true for Spanish and the majority of the other European languages, as well as for the native tongues of the multitudes of immigrants who speak languages such as Thai or Cantonese or Farsi.

The Native American languages don't have this advantage, nor do they live on in collections of books and audiotapes and movies. Unlike the case of the Hebrew language of Israel, no ancient books upon which a modern revival could be based exist in these languages. There is literally nowhere on earth the child of Quapaw ancestry can go to find out how his or her ancestors spoke: Quapaw is lost forever. If the hypothesis is correct, if losing a language does mean losing a culture, then the Quapaw culture is gone forever.

This may or may not trouble you. If you are a supporter of the "survival of the fittest" ethic, you may consider it acceptable. It troubles me, and like many others I view it as a tragedy almost beyond comprehension; I know no coherent way to decide what "fittest" means in the context of languages and cultures. But how would we prevent the extinction of languages?

Think what it would cost to train teachers and prepare teaching materials and establish curriculums in nearly 200 different languages, even if roughly adequate grammatical descriptions were already available for them. How would we find those funds in a country where the struggle for education dollars—for instruction in English, not to mention Spanish and Cambodian and Chinese and a host of other languages spoken by immigrant schoolchildren—is already deep and bitter?

One obvious solution is to say that clearly we can't save all the languages but that we must save some of them. (The "fittest" ones, presumably.) But how do we choose which ones to save? What individual or agency would we give the authority to make those choices, and on what basis? We can never be sure in advance which rain forest plant or animal will turn out to be the one that cures a terrible disease, and our scientific foundations for choosing which languages to protect are far shakier than those provided for snail-darters and louseworts by the biological sciences.

Linguist Einar Haugen (Vines 1996, page 27) argues for "thoughtfully planned bilingualism." He wants every single one of us to have a "native, homely, familiar everyday language in which we can live and love" plus another language "of wider communication."(See Chapter 7 for a further discussion of this possibility.) But however attractive that idea might be, the practical problems involved in putting it into practice are so formidable that it's hard even to know where we would begin.

The English Only contingent in Congress seemed for a while to have decided that the proper attitude toward all languages spoken in the United States other than English was "Since we can't save them all, we won't save any of them—that's fair!" They had a minor change of heart, later; they did come up with $1 million specifically for preservation of the Native American languages, after the case for their special status was presented.

But the problem still remains: How do we choose which of the Native American languages to give funds to, and on what basis? Who gets the power to make those decisions? One million dollars is a very small sum to divide among nearly 200 language preservation programs.

"Right! And the Sooner That Happens, the Better!"

It is absolutely wrong and against American concept to have a bilingual education program that is now openly, admittedly dedicated to preserving the native language and never getting them adequate in English so they can go out into the job market.

(President Ronald Reagan, in Hakuta 1986, page 207)

Winona Simms, a Creek-Yuchi Indian and director of an American Indian program at Stanford, recently told us, "You would think that being bilingual was the equivalent of being learning-disabled." This logic also led an Amarillo judge to tell Martha Loriana two years ago that speaking Spanish to her daughter bordered on child abuse.

(Rodriguez and Gonzales 1997)

Skeptics answer with … the Tower of Babel. Just as conflicting languages kept the tower's builders from reaching heaven, they say, language preservation only keeps us further from peace and mutual understanding. With ethnic wars raging across the planet, shouldn't we seek a common tongue to bind us rather than reviving languages to further divide us?

(Bilger 1994, page 38)

That last quotation is particularly instructive in the context of the ethnic wars that have been ravaging the Balkans. Historian Ivo Banac writes (in Woodward 1996), "The fighting may be over, but the

The Link Between Language and Culture successor states of Yugoslavia are waging new wars over words … The whole point is to create new differences between Croatia and (Serb-dominated rump) Yugoslavia so that communication between the two is more complicated and the idea of separate identities strengthened … Croatian authorities are aggressively 'purifying' their country's language by substituting words deemed to be foreign with Croatian words."

Like those who want to preserve and protect languages, those who respond this way believe that the hypothesis itself is correct, and that cultures are lost when languages are lost. But they don't consider that a tragedy or something to be struggled against. On the contrary! Here's Republican Congressman Randy Cunningham of California arguing during a debate for the English Only legislation (a debate that Rep. Pat Williams said was "the most maddening" he'd sat through in his eighteen years in Congress):

If you want to keep your people in a barrio, if you want to keep them restricted in little tight communities so only you can communicate with them, and we can't in English, then be my guest. But we're going to empower them to get out of that so they can have a piece of the American dream.

(Quoted in Schmitt 1996)

Republican Robert Livingston of Louisiana directed his colleagues' attention to the separatist movement in Quebec, predicting that Canada won't last another decade as one country, and said, "We already went through one great civil war. We don't want to go through another" (also in Schmitt 1996).

Although all observers may be confronted by the same physical evidence in the form of experiential data and although they may he capable of "externally similar acts of observation," a person's "picture of the universe" or "view of the world" differs as a function of the particular language or languages that person knows.

(Lee 1996, page 87)

First you have to claw your way through the linguistic thicket created by the academic register in which that quotation is written. Why is it written like that? One of the rules of the Academic Regalian register is that the more you expect other academics to be opposed to what you're saying or writing, the more extreme your use of the register has to be. This is unfortunate, because controversial subjects are also subjects about which it's important to be as clear as possible. But if the academic game is the game you're playing, clarity has to be sacrificed to this linguistic dominance display.

The translation process introduces a delay, certainly; but when you get to the end of it you will realize that you've come upon a concept so interesting that it grips the mind and won't let go. It's called "the linguistic relativity hypothesis" (also "the Sapir-Whorf hypothesis" and "the Whorf/Whorfian hypothesis"). Lee's quotation says that the way human beings perceive the world around them varies with the languages they know—even though they perceive the same things, in the same manner, using the same physical and mental "equipment." This will strike you either as common sense or nonsense, depending on your own personal convictions about the power of language.

Linguist Dan Moonhawk Alford offers the example of a Cheyenne parent who is sitting with a child on his knee when a ball suddenly bounces across the floor. An English-speaking parent would say to the child, "Look! Ball!" The Cheyenne parent, Alford tells us, would say, "Look! Bouncing!" (in DellaFlora 1998). Both parents are perceiving the same stimulus, with the same sensory equipment. However, the way they express the perception—which tells the child what it's important to pay attention to—is not the same.

In 1958 Ace Books published Jack Vance's science fiction novel The Languages of Pao, in which a government tailored its population for specific roles in adulthood by controlling the language each person learned in infancy. One group of infants learned a native tongue that fitted them for life in the military, another group learned a language designed for business and trade, and so on. This worked very well—until the government was overthrown by a population that had secretly learned more than one of these designer tongues natively and had therefore grown up with more flexible perceptions of reality.

The Languages of Pao is a straightforward presentation of a fictional world having these three characteristics:

1. The linguistic relativity hypothesis is valid and true.
2. The hypothesis can be systematically applied to human life.
3. The government has the power and resources to carry out that systematic application as a national policy, beginning with its population of infants.

Whether (a) and (b) are true in this nonfictional world that we all live in is a matter for fierce dispute; I am firmly convinced that they are. Because we know of no government in human history that has met the specifications for (c), it's difficult even to speculate about the truth or falsity of that proposition; in the novel, its truth is presupposed.

Suppose that the linguistic relativity hypothesis is true in the real world. Suppose we could prove that languages actually do have significant control over the worldviews of their speakers. What would that mean?

There are two major possibilities. If we assume that a multiplicity of worldviews causes disunity in a society, it would mean that we could make a strong case for the proposition that multilingualism at a national level is dangerous and should be discouraged, even forbidden. (This idea is in many ways at the heart of the various "English Only" movements in the United States, although many of their proponents are unfamiliar with the linguistic relativity hypothesis.) On the other hand, we could assume that the more worldviews people have, the better; in that case, the validity of the linguistic relativity hypothesis would let us make a case for strong support of multilingualism. (This would be my personal choice.)

In addition, we would be able to make the case that letting a government decide which language or languages people will learn is dangerous, since that would mean that their government dictated their view of the world for them. Clearly, both life and government policy would be less complicated if we could be certain that the hypothesis is false.

The linguistic relativity hypothesis (LRH, from now on) is a source of controversy in many fields, including at least linguistics, anthropology and ethnology, philosophy, religion, political science, and education. It's not a new idea; it goes back at least as far as Giambattista Vico, who lived from 1668 to 1774. Today it's associated primarily with the work of linguists Edward Sapir and Benjamin Lee Whorf, especially Whorf. Linguists accuse one another of being "Whorfians," acknowledge (like me) that they are Whorfians, or deny ever having been Whorfians, all with an astonishing degree of passion.

The LRH not only has more than one name, it has more than one formulation. First, there is what's known as "the strong version," which says explicitly that human perceptions are controlled by human languages, as assumed in The Languages of Pao. Oddly enough, this version (often called "linguistic determinism," as if yet another name were needed!) is itself science fiction. No scholar (including Sapir and Whorf) has ever suggested—much less openly claimed—that any such hypothesis is valid. Think about it for a moment. It would mean that if your language has no word for chair, and I set a chair down in front of you, you wouldn't be able to see the chair; it would be impossible for you to perceive it in any way. If you tripped over it and fell down, you would presumably be baffled by that experience. This idea is nonsense; no question about it. Why, then, does it exist? We'll come back to that in a minute.

The other formulation, called "the weak version," makes no such absurd claims. It says that human perceptions of reality are structured and constrained—not controlled, but structured and constrained—by human languages, in interesting and significant ways. As linguist Dan Slobin (Slobin 1998, page 1) has said, Steve Pinker may be right in claiming that we lack solid scientific evidence that languages have dramatic effects on human thought, but science should not take that as an excuse to ignore the nondramatic ones. For all we know, the reason we find no dramatic effects is nothing more than our lack of skill in applying the LRH to language. Unlike the strong version, the weak version can't just be labeled nonsense and pitched over the side, as we will see in this and later chapters. If it's true, it matters, because it means that language has a degree and kind of power over human thought that many people (including many scholars and scientists) find offensive and/or alarming.

Which brings us to the reason why instructors continue to teach their students that "the linguistic relativity hypothesis has a strong version and a weak version." It has become a traditional tactic to refer to the overwhelming evidence against the fantasy strong version when arguing that the weak version is false. For that purpose (and so far as I can see, only for that purpose), there has to be a strong version. The very fact that any behavior as silly as this goes on among scholars and scientists demonstrates how important the LRH is; if the hypothesis were trivial, the silly behavior would have died out long ago. Throughout this book, therefore, we're going to discuss the so-called weak version, to find out why it is so important and what its implications are in the context of the many controversies about multilingualism and language policies. I will specify here—for once and for all, so I don't have to keep saying it—that unless I tell you differently on the spot, when I refer to the LRH from now on in this book I am referring only to the weak version.

It's encouraging to be able to report that the policy of ridiculing the LRH seems to be beginning at long last to weaken. The September 1998 issue of Language (the most important journal in linguistics) not only had a positive review of Penny Lee's book defending the LRH, it also contained an article presenting some compelling new evidence for it. Eric Pederson and his associates conducted extensive research over thirteen different languages from ten different language families. It has always been assumed that human beings structure their basic perceptions of space—such as the locations of up, down, left, right, in front, and behind—by reference to their own human bodies, and that this is universal for humankind. Pederson's research tells us that when you move beyond the languages most commonly used in the Western world, you find that it's not universal after all. (For details, see Pederson et al. 1998. See also Slobin 1998.)

The Linguistic Relativity Hypothesis in Action

The LRH, if it's valid, can operate either between different languages (the traditional formulation) or between varieties within a single language. Let's look at examples of both kinds, as well as examples where the two are intermingled, to make that concept more clear.

Who's in Charge, the Human or the Horse?

Suppose you wanted the following English sentence—"I was riding a horse"—translated into Navajo. Monolingual people who are inexperienced with translation often assume that you translate from Language X to Language Y in three stages: You find the words in Language Y that have the same meaning as the words in Language X, you put them in the order required by Language Y, and then you tidy up any leftover details. That can be accurate, especially when you're translating between two closely related languages, but most of the time it won't work. To get accurate translations you have to set up a scenario in which a speaker of Language X would say whatever it is you want translated, and then find out what a speaker of Language Y would say in the context of that same scenario. Let's say that our scenario goes like this:

John has said to Bill, "Hey, I tried to call you a couple of times yesterday afternoon, but you never answered your phone. What were you doing?" John and Bill are good friends, and Bill is more than willing to answer the question; he says, "I was riding a horse."

We could describe this scenario to a native speaker of Navajo and ask, "If Bill and John spoke Navajo instead of English, what would Bill say in order to answer John's question?" At which point a long discussion about the Navajo verb would suddenly become necessary. The Navajo speaker would have to know many more details about this "riding" before the correct verb form could be chosen. Was Bill riding this horse with a destination in mind, and if so, was he going to stay at the destination, or was he going to go there and then come back again? Or was Bill just riding around aimlessly? Was Bill riding this horse as a one-time event or is it something he does frequently—and if it's frequent, is there a pattern to the frequency? Lots more information would be needed; many pages of this book would be required if I wanted to make it all even roughly clear.

We're going to set all that aside and focus on the way that "I" and "a horse" would be translated. This will be more than enough to make the necessary point.

Classroom-English GrammarSpeak says (awkwardly) that "I" was doing the riding, and that "the horse" is what the riding was being done to; the horse is called the object (usually, the direct object) of the action. According to my Navajo consultants, the Navajo sentence isn't like that. In Navajo, no creature or thing is having anything whatever done to him or her or it. The Navajo sentence is like "Mary was dancing with Tom," where Mary and Tom are dancing together and neither of them is the "object" of the action. Furthermore, the most likely Navajo verb in this context would be one used only to refer to the acts of animals (or to make jokes about people on that basis). The horse, in the most likely Navajo equivalent for "I was riding a horse," is said to have been "animaling-about with me"; the other likely possibilities would mean something along the lines of "a horse and I were moving about together." You don't have to be fluent in Navajo, or a professional translator, to realize that these sentences mean something rather different from what "I was riding a horse" means; nevertheless, they are the correct translations for this scenario. They are what the native speaker of Navajo would say in the situation in which the native speaker of English would say, "I was riding a horse."

Chinese does this the way English does it; the Japanese equivalent means that "on a horse" is where "I" was, not that anything was being done to the horse. Hopi has the human being doing something with the horse, but it's not like the Navajo sentences, in which the horse and the person are associated; in Hopi, it's the human being who is in charge. Hopi has the human riding around "with a horse" in exactly the same way it would have a human eating soup "with a spoon." The horse, in Hopi, is the instrument with which the action is done.

We don't need to concern ourselves with the grammatical details of all these data, interesting though they are. The question that matters for our purposes is:

Do these different ways of giving a spoken shape to an experience mean that the human beings involved think about the experience, and about the elements of the experience, in significantly different ways?

A Whorfian would suggest that someone a horse "animals-about" with might, as a result, have an attitude toward horses and horseback riding that would be different from the attitude of a person who uses a horse to go about with, or a person who does something called "riding" to a horse. I'm saying "might" here because you can't come to firm conclusions about this on the basis of just one example. But the single example, to a Whorfian, means that the conclusion is sufficiently likely to be worth further investigation. An anti-Whorfian would say without hesitation that such a conclusion is silly rampant romanticism with no foundation in reality, and that classifying horses as objects or instruments has nothing to do with how horses are perceived. An anti-Whorfian would remind us that speakers of English talk about hearts of celery and hearts of artichokes without the idea of vegetable heart attacks ever crossing their minds.

English speakers tell a restless child, "Be good!" The French, in exactly the same situation, tell the child to "Be wise!" Could that make any difference in the child's perceptions of the world around him or her? Could it make a difference in what the child concludes deserves attention and what can safely be ignored? Could it make a difference in the child's behavior in the real world? The Whorfian would say that it very well might; the anti-Whorfian would say that it couldn't possibly. Let's look at a real-world example.

One of the severe health problems that India struggles with results from the fact that the Ganges River is used not only for bathing and washing and religious rituals of various kinds but also for disposal of waste and disposal of corpses. That seems straightforward enough. But health workers' attempts to explain this to people and help with the problem have run into difficulties, even when (as is often the case in India) fluent English is spoken by everyone involved. Why? Because devout Hindus consider the Ganges holy (a belief which doesn't have its roots in English). How could something holy be polluted? That's an impossible contradiction. Furthermore, it's a dreadful insult. As Fran Peavey writes, "In the context of India, that would be like my saying 'Your mother is a whore' to a Westerner" (Peavey 1995, page 49).

The source of this quandary is in the word "pollution" and its associated meaning and connotations. When Peavey discussed the situation with the people along the river, she found it hard going, but she persevered. After many long conversations and much effort to explain and to understand on both sides, what she would hear from them was this: "The river is holy, but she is not pure. We are not taking care of her the way she needs us to." That is, the river is not "polluted" but "neglected," and it is the people themselves who are neglectful. And then, Peavey tells us:

The funny thing is that, after hearing this reply, I started to think less in terms of "pollution" and more in terms of "people not taking care of the river." This was an important change of perspective. "Pollution" is an abstraction that does not address the responsibility of the people who are making the mess. "Pollution" focuses attention solely on the river. It is almost as if the river is to blame for being polluted!

(Peavey 1995, page 40)

Now health workers have begun saying—in English—"The Ganges is suffering" rather than "The Ganges is polluted" (Ward 1997, page 51). This despite the fact that in English rivers cannot "suffer." With the new way of naming the situation, it has become possible to talk about it, and about potential remedies, without giving or taking offense. The facts have not changed; all the people involved are still perceiving the same event with the same eyes and minds; but the outcome in the real world is very different than it was before. As Howard Rheingold puts it:

Finding a name for something is a way of conjuring its existence, or making it possible for people to see a pattern where they didn't see anything before … We think and behave the way we do in large part because we have words that make these thoughts and behaviors possible, acceptable, and useful.

(Rheingold 1988, pages 3-4)

Alford puts it even more strongly:

People who are monolinguals and know and think in only one language are handicapped. Because they don't have another worldview to shift to, to look at the same problem and see if they come up with a different answer.

(in DellaFlora 1998)

Who's in Charge, the Human or the Language?

People who are willing to accept the idea that two languages (especially two very different languages) might cause their speakers to perceive the world in different ways will still often balk at the idea of such effects within a single language. "If the linguistic relativity hypothesis has any validity at all," they'll say, "and mind you, I'm not saying it does, I'm just saying I'm willing to consider the possibility—then it has to be valid cross-linguistically. Inside one language, it's nonsense!"

I disagree. For example, let's consider the case of Carol Cohn, a woman who, although strongly opposed to nuclear war, became a visiting scholar in a nuclear strategic studies program at a major American university. The nuclear situation at that time (in 1984 and 1985) struck her as "so dangerous and irrational that one is tempted to explain it by positing either insanity or evil in our decision makers." Her goal in joining the program was "to gain a better understanding of how sane men of goodwill could think and act in ways that lead to what appear to be extremely irrational and immoral results" (Cohn 1987, page 17).

In order to fill her role as visiting scholar, Cohn attended lectures, talked with colleagues, interviewed graduate students, and did all the other academic tasks expected of her. She says that she started out obsessed by the question, "How can they think this way??" But as she became more skilled in using the language of the program and more at ease with it, she found herself facing a different question: "How can I think this way?"

The English vocabulary of warfare (both nuclear and nonnuclear) is curious and ingenious. We keep our missiles in "silos"; the area on a Trident submarine where the silos are located is called "the Christmas tree farm." We "marry up" weapons systems when we combine them. One nuclear attack model is called "the cookie cutter." We use "clean bombs" to carry out "surgical" (that is, surgically clean) strikes. When we're not using domestic images of family and kitchen and farm, we still use terminology clean enough to eat off of. When we talk about human death resulting from warfare on the evening news, however bloody and messy it may have been, we don't talk about "dismembering and decapitating people" or "burning people into ashes and charred fragments" or anything of that kind. We don't even call it "killing"; we call it "collateral damage." When we refer to the deaths of civilians we ordinarily make those who died the do-ers rather than the done-to. We don't say "We killed three hundred civilians in yesterday's bombing," we say "Three hundred civilians died during yesterday's bombing." As if, like the polluting Ganges, they did it all on their own.

Cohn wasn't able to function in the nuclear studies program without speaking its language. No one would accept her as a valid nuclear scholar, or take what she said seriously, or communicate with her in any of the ways her role required, unless she used English WarSpeak or NuclearSpeak. And as time went by, even though she remained as solidly opposed to nuclear warfare as she had ever been, she made a disturbing discovery. "No matter how firm my commitment to staying aware of the bloody reality behind the words," she writes, "I found that I could not keep human lives as my reference point. I found I could go for days speaking about nuclear weapons, without once thinking about the people who would be incinerated by them" (ibid, page 22, emphasis mine). It will surely be obvious that people who can be put into this frame of mind are more likely to function efficiently as military planners and personnel than are people who keep thinking about charred corpses.

To maintain the view of reality that she'd had when she entered the program, in which she perceived nuclear war as a matter of blood and guts and agony, Cohn had to constantly remind herself of those perceptions. And she couldn't do that while she was actually involved in language interactions in the program, of course; she had to do it on the side, on her own. She found it difficult. She had to work at it.

So? It's an interesting little anecdote, but does it have any relevance in the real world? Think about it, please. Dr. Cohn was able to come out of this experience with her worldview intact. But she was "immersed" in the nuclear warfare language environment only for a short time, and she went into it as a mature adult with opinions already solidly formed. What if she—like the students in our elementary schools, for whom we do choose the language of instruction—had been a small child? The Languages of Pao proposes that you could raise adults who would be superbly fit to do battle if you started them out in childhood immersed in a warrior language; their adult worldview would be a warrior's worldview, and everything about combat would make sense to them and strike them as natural. What sort of grownups would we get, do you think, if we selected a group of children and did our best to carry on their education in NuclearSpeak or one of the other varieties of warfare-English?

Another area within languages in which we see Whorfian effects is language tightly linked to sexual gender. Much anger and many confrontations revolve around "inclusive language" in Christianity and Judaism, for instance. Some people consider the revision of Judeo-Christian scriptures and liturgy so that it more explicitly includes female human beings a tremendous improvement; some consider it nearly blasphemous; and some consider it merely silly. For the sake of discussion, let's assume that inclusive language is a good idea. To make religious language inclusive of women, all you do is replace masculine items with feminine or nongendered ones. So "sons of God" become "children of God" and "mankind" becomes "humankind"; an endless series of "he" and "him" and "his" and "himself" items is dealt with by making the verbs plural wherever possible, so that the genderfree "they" and "them" and "their" and "themselves" can be used instead. A simple process, right? And, since everyone agrees that God does not have any human sexual gender, entirely mechanical. But look, please at this example:

The Lady is my shepherdess; I shall not want.

Do you see her, in your mind's eye? With her blowing skirts and froth of lacy petticoats, and her ruffled pantalettes showing, and her sweet little crook with its pretty bow tied on it? It's Little Bo-Peep. She won't do; this is a very bad "translation" indeed. (We'll return to the inclusive language debate in Chapter 6.)

A similar problem caused our English-speaking news commentators during the Gulf War to start out with "Today our boys and—," come to an abrupt halt, and switch immediately to "Today our men and women in the Gulf [did something or other]." "Our boys and girls in the Gulf" would have sounded as if the military personnel were all sitting in sandpiles playing with little plastic buckets and trowels, not making war; the childish connotations of the word "girl" leak onto "boy" when the two are linked, causing semantic contamination. (It's semantic contamination that makes us accept "He's a dashing bachelor" but balk at "She's a dashing spinster," even though we're taught that "spinster" and "bachelor" differ in meaning only by gender.)

If the LRH isn't valid, effects like these should be extremely rare and always brief; that is, they should be something that is noticed the first time or two that the sequence of language occurs but is forgotten thereafter. That's not what happens.

Martha Heyneman writes, "The mother tongue … orders the savage landscape of our raw sensory impressions, domesticating them, molding them into a world we can step out into with relative confidence, the world of common sense" (Heyneman 1992, page 4). If that is so, and if the LRH is valid, then the decision as to which language(s)—and which varieties of languages—shall be the native tongue(s) of those who will grow up to run a country becomes extremely important. It becomes important to decide which foreign languages, if any—and which additional varieties of native languages, if any—people should be allowed to learn. (We've seen a passionate and confused example of the latter issue in recent years, in the controversy over "Ebonics.") If the LRH is valid, then certain languages, and certain varieties within languages, will tend to mold certain sorts of people. Learning certain languages and varieties may have significant effects on adults, even though they are already molded. It's easy to see why scholars and scientists and policymakers—as well as producers of mass media, such as violent films and computer games—might greatly prefer that this should not be true.

Steven Pinker of MIT is one of the world's most prominent linguists; he is that very rare bird, a linguist celebrity. In 1994, in his best-selling book The Language Instinct, he made it clear that he has no patience with the LRH; he refers to it, as is typical, in its strong form, which he calls "linguistic determinism."

"It is wrong," he says on page 57, "all wrong." He tells us on page 58 that "Once a euphemism is pointed out, people are not so brainwashed that they have trouble understanding the deception."

Let's consider that. There is a form of birth abnormality associated with significant mental retardation and a particular sort of facial appearance which includes slanted eyes. Children born with this abnormality used to be called "Mongolian idiots." Today we have a euphemism that we use instead: We say that these children "have Down's syndrome." The theoretical result of Tinker's position would be a claim that the two dialogues below are synonymous and interchangeable in every way except stylistically:

1. TOM: What's the matter with that kid?
   BILL: She has Down's syndrome.
   TOM: I see. That's a shame.
2. TOM: What's the matter with that kid?
   BILL: She's a Mongolian idiot.
   TOM: I see. That's a shame.

I think Pinker is "wrong, all wrong" about this. I think that saying these children "have Down's syndrome" instead of "are Mongolian idiots" is much more than "style" or "being polite." It makes a tremendous difference in the way that people perceive the children and behave toward them, as well as in the way the youngsters perceive themselves. As Carol Cohn discovered, human beings, even very intelligent and sophisticated human beings, are not impervious to euphemisms simply because they understand that that's what they're dealing with.

I agree with Pinker that these things shouldn't matter. In a perfectly logical world, what a medical condition is called should have nothing at all to do with the way those who have it are perceived and treated, by themselves or by others. But we know, from large quantities of excellent research, that people don't consistently base their decisions on logic; two minutes of thinking about their behavior with seat belts will more than make that case. However illogical it may be, attitudes and behaviors in the real world toward "a Mongolian idiot" will unquestionably be different from attitudes and behavior toward "a person who has Down's syndrome." That, in my opinion, is one of the powers that language has; it's not trivial, and it should be used with exquisite care. When I imagine someone telling another person, "You know, that Down's syndrome thing just means that the kid is a Mongolian idiot"—Pinker's "pointing out a euphemism"—I dislike that imaginary dispenser of information intensely. I think that he or she is doing real harm.

However, although we're nearing the end of this chapter's discussion of the link between language and reality, we are still floundering about, mired in scholarly squabbling. To find out whether the hypothesis that sugar melts in water is valid, all I have to do is drop a sugar cube in a glass of water and pay attention to what happens next. The things I would need in order to prove or disprove the linguistic relativity hypothesis aren't like that; they're located somewhere inside the human mind, where we can get at them only clumsily and indirectly. Although people are quick to say that the LRH is right or wrong, and quick to accuse those who take the opposite position of shortcomings of all kinds, they—and I—are only offering you opinions. The jury in this case is still out. The linguistic relativity hypothesis has neither been proved nor disproved. Strong evidence exists for both sides, and you're free to decide for yourself whether you consider it valid, in a way that you cannot'decide for yourself about that cube of sugar. For a thorough account of the entire controversy about the LRH—with careful reviews of the arguments, the squabbling, and the evidence both for and against, to help you make your decision—I strongly recommend that you read Penny Lee's The Whorf Theory Complex: A Critical Reconstruction, in spite of its academic language.

With all that said, we can go on now to consider a phenomenon that dramatically illustrates the effect language can have on perceptions and behavior.

Case Study: The Japanese Water Babies

I first came across the water babies concept when I read this sequence in Kittredge Cherry's elegantly titled and irresistibly interesting book Womansword: What Japanese Words Say About Women:

Thousands of Japanese women visit temples every year to bring baby-clothes, cookies, toys, and other offerings to their mizugo … The mizugo is a fetus, removed from the watery warmth of the womb by nature or abortionist before it sees the world. "To make a water-baby" (mizugo ni suru) means to have an abortion.

(Cherry 1987, page 79)

I was literally thunderstruck—an old-fashioned word, but the exact and perfect one in this case—by Cherry's account. But I was also dubious. I thought that surely this was either folklore or a description of a custom from long ago. I tried to imagine two American women talking to one another about what they planned to buy as a gift for their aborted fetuses this year, and found the image impossible. So I proceeded to check it out.

With the help of respondent Fran Stallings (English/Japanese) and her friend and colleague Hiroko Fujita (Japanese/English), I learned that the water babies not only aren't folklore but are very much a part of contemporary Japanese life. Hiroko Fujita was kind enough to send me a postcard of one of the most famous temples (the Temple of Jizo on the Mountain of the Purple Cloud), showing row after row after endless row of the little statues that Japanese families put there to honor their water babies, both those that result from miscarriage and those that are the result of abortion. Hiroko Fujita sent pages of detailed information; and Dr. Stallings directed me to a book by William R. LaFleur, Liquid Life: Abortion and Buddhism in Japan, where I found a thorough account and discussion.

LaFleur explains that it's not just mothers who go to the temples. Children—the water babies' siblings—go along; often whole families join in these visits. The statues memorializing the water babies are images of Jizo (a guardian of children, in the sense that Catholics might speak of a patron saint). They are all almost exactly alike, carved of stone, roughly two feet high, with bald heads and closed eyes. They are stone metaphors: They look like both tiny Buddhist monks and tiny children, at the same time. Over their carved monks' robes they wear the large red bibs that Japanese infants and toddlers wear, tied around their necks by the families who visit them. Often amulets signed by family members are hung in the temples, inscribed with messages to the water baby expressing the family's debt of gratitude. There is no attempt to hide what has happened or to make it disappear from the family memory. Rather:

Loving attention to the dead is shown by washing down the memorial image—an ancient Buddhist practice—providing fresh flowers, and bringing the occasional new toy or garment. These activities and the recitation of simple prayers are expected. But beyond these there is the sense of an active communication, emotional if not verbal, between the living family and the departed child.

(LaFleur 1992, page 11)

Notice, please, that the word LaFleur ends with above is "child." We've come upon another lexical gap, of a somewhat different kind. English has no word or phrase that would mean what we have been translating into English as "water baby." If we were asked what an American English speaker would say in the same situation, silence would be our probable answer, because American English has no vocabulary we can easily or conveniently use. In Womansword (page 80), Kittredge Cherry provides the linguistic spotlight that shows us the gap when (struggling with the clash of languages, as I am also struggling) she writes that the Japanese "do not always draw a clear distinction between a fetus and a baby."

Cherry writes as if American English (AE) does make a clear distinction in this matter, and as if speakers of AE can identify a clear consensus about that distinction—but of course that's not true. For sure, a human being who has gone through the full experience of childbirth is a child. But our entire national debate about abortion, which now is referred to far and wide as our "second Civil War," revolves around the question of when, precisely, the united sperm and egg are a fetus and when they are a baby. If we had "a clear distinction" in that regard, we would perhaps be able to sit down as a nation and make the necessary decisions; we might not still be in turmoil. LaFleur tells us (page 11) that the Japanese "tend to avoid terms like 'unwanted pregnancy' or 'fetal tissue.' That which develops in the uterus is often referred to as a 'child'—even when there are plans to abort it." The power of language is such that it would be an extremely rare mother and native speaker of English who could bring herself to say, "Tuesday afternoon at three-thirty I'm going to abort this child."

The visits to the temple and its cemetery are not the only acknowledgments of the water babies in Japanese society. Many Japanese familes have a household shrine. The aborted child (baby? fetus? fetal tissue? product of conception?), although not an ancestor, is often honored at the shrine and "is treated as an ancestor and will be remembered along with them at the Buddhist altar in those households that have them" (ibid., page 148).

The real-world difference that makes this example so important is (in painful contrast to the the terms we have available in English for discussing it) easily stated:

The water babies are not part of this world, but they remain part of the family; they have a distinct role within the family.

Now you might object that this is not a difference of language but "a cultural difference," or "a religious difference." Many Japanese believe in reincarnation, and believe that the fetus (child? baby? product of conception?) will have another chance, even many more chances, at life on this earth. Certainly, that is a difference of religion and of culture. But language is what we use to express cultural and religious differences, and language is the mechanism we use to transmit information about those differences to others, including our children and our grandchildren. It will be clear to you, just from the problems I've been having writing coherent sentences that explain this Japanese concept in English words, that it makes a real-world difference whether you call this entity a fetus or a baby or something else entirely. As with the linguistic strategies of our own WarSpeak, it makes a difference that the process of government certification for abortionists in Japan makes them members of the Motherhood Protection Association. It's not just a matter of style or manners; it's not, as Pinker has suggested, just a matter of people's susceptibilities to "brainwashing." Women in the United States, not just Japanese women, demonstrate that every day.

Millions of English-speaking women (myself included) have said, "I felt the baby move today"; I have never heard, or heard of, a woman saying, "I felt the fetus move today." No one says to a pregnant woman, "When is your fetus due?" or "What are you going to name your fetus?" No woman says, "I had an ultrasound today, and I'm so pleased—my fetus is a little girl!" I have never heard any woman say "my fetus" (although I have heard doctors say "your fetus" to women); always, for the woman, it is "the" fetus.

The word or phrase we choose when we refer to the "product of conception" commits us, as the linguistic relativity hypothesis would predict, to certain attitudes and certain potential actions—not only with regard to the "product" but with regard to its mother and father (notice how hard that is to process in English!) and other aspects of the situation. Whether we choose a single word or phrase to refer to both abortion and miscarriage, as many languages do, or use two separate words as English does, commits us to certain attitudes and potential actions. Whether the naming we do makes it possible to give the "product" in question a role within the family or not drastically changes our commitments.

The question that immediately comes to mind when we think about this case study in the context of multilingualism is:

What effect does the water baby concept have on attitudes toward and feelings about abortion in someone who has both Japanese and English as native languages?

The anti-Whorfians would say that the question is irrelevant, that such attitudes and feelings are based not on language but on logical consideration by the individual, or on socialization, or on a combination of the two.

I disagree. I don't speak one word of Japanese; nevertheless, just learning about the water babies by reading accounts in English had immediate and significant effects on my own feelings and attitudes. I thought of the huge boost that the United States greeting card and holiday industries would get if we were to add the water baby cultural construct to our American culture; I also thought what a comfort it might be to many women and to their families if that happened. My perception of abortion (and miscarriage) has been permanently changed by my encounter with this phenomenon in language, as has my perception of "a family"; the concept of the water baby presents an entirely new option in the world that had been invisible to me previously. It seems to me impossible that anyone who speaks both languages natively would not suffer to some degree the discomfort of the psyche that Leon Festinger named "cognitive dissonance."

We need to know how Japanese/English speakers resolve conflicts of this kind. Do they choose one of the alternative worldviews over the other? Do they negotiate some sort of compromise worldview that takes features from each, so that they "average out" in some way? Do they reject them both and create a third alternative that is unique to those with their language competencies? Is the conflict impossible to resolve, forcing them to maintain both worldviews and switch back and forth between them, as Carol Cohn had to do in her role as nuclear studies scholar? Can they maintain both world-views without discomfort, in an amazing display of what it means to be broad-minded? Is there yet another alternative that hasn't occurred to me, perhaps because my language makes it difficult to perceive it? How does this all work out? What are its implications in a global society? What happens, to take just one possible example of many, when English-speaking diplomats and officials (who rarely speak Japanese) carry on "family planning" conferences with their Japanese counterparts (who almost without exception speak English)?

This is the paradigmatic problem, the pattern question at the heart of our investigation. Suppose it's true that languages structure and constrain our worldviews. Then how do multilinguals deal with conflict between two (or more) worldviews, especially on issues that are highly charged emotionally? What implications do multilinguals' strategies and solutions in such situations have for the society and culture in which they live—or for the world at large? If it's not true, of course, matters are much simpler, and decisions about language policy are far more easily and conveniently made.

Advocates of measures such as "English Only" (when not basing their support totally on financial or political factors) are making a set of broad assumptions. Using the Japanese water baby concept as our framing example, we can describe what they mean when they say that "English Only" is necessary to "becoming an American." (That expression is used constantly in English, even by those who realize how inaccurate it is given all the peoples of the Americas who aren't Anglos. In the absence of something like "Unitedstatesers" to fill the lexical gap, it's difficult to get by without it.) The "English Only" contingent mean that their policies will bring about a sequence of events like the following :

• A little boy arrives at an American school as an immigrant, speaking only Japanese.
• He learns English and begins relying on English in his daily life.
• As a result, he undergoes a cognitive transformation.
• One effect of that transformation can be summarized roughly as "When I came to the United States I thought there could be an honorable role for an aborted or miscarried fetus in the family that its mother belongs to, but now that I speak English I know better."
• A transformation of this same kind occurs with regard to every cultural concept for which a significant conflict exists between American English and Japanese.

If both languages are maintained in the youngster's life, as in multilingualism, how can that cognitive transformation happen?

Many Native Americans would chime in right here and say, "That is precisely our point. That is why we must not let our languages die!" And many immigrants, even those whose languages are safe because they're spoken elsewhere in the world, would say that this is why they must insist that their children maintain both (or all) of their languages.

The real problem is that we aren't paying proper attention to what we're doing. It's not just that we don't know with certainty whether giving up a language means giving up a culture or whether giving up a language means giving up a worldview. It's not just that we don't know the cognitive result when the worldviews of a multilingual person's languages are in severe conflict as they are for English and Japanese in the context of abortion, and we don't know what strategies are used to deal with such conflicts. The real problem is that we aren't taking our ignorance in these matters into consideration in making language policy decisions.

My personal conviction is that the link between language and culture is so deep and so strong that the loss of a language inevitably means loss of the culture and its worldview, a worldview unlike any other. But that's a conviction which I realize is based on inadequate information. In the scientific sense of knowing, we don't know, and so we try to simply muddle through somehow, with the sort of consequences that are typical of muddling through. It would be a tremendous improvement if everyone involved were at least aware of what the unresolved questions are and what evidence we do have. The evidence for the strong link between language and our perceptions of reality is so compelling that we must at minimum begin a public debate; a debate limited to scholars and experts is grossly inadequate.

It may be that we'll never be able to resolve the matter with real certainty. We can't impose languages on people in the way the government imposed them in The Languages of Pao. We can't remove languages from brains and isolate them in laboratories for study. The tools we have for observing brains busy "languaging," such as the various PET and MRI and fMRI brain scanners, are difficult and expensive to use and restricted to the medical environment; experimental results therefore tend to be based on data from perhaps six to ten brains. Results like that can teach us a lot, but we can't rely on them to prove or disprove our hypotheses. A dozen brains might give us one clear result, but we know that if we looked at a thirteenth or sixteenth or fiftieth brain it might show us something entirely different. In most sciences results are considered respectable and reliable when they're based on hundreds of carefully examined examples, not a dozen.

What we can do, however, what I believe we're obligated to do—is to try to bring the sets of assumptions that people are relying on for language policy decisions to conscious awareness. We need to state those assumptions as clearly and explicitly as possible. We need to make them widely known, the way it's widely known that sugar cubes melt in water. We need to examine the evidence for and against those sets of assumptions, however limited it may be, so that there can be rational public debate about their accuracy and so that the links between assumptions and policy can be traced. This wouldn't be a solution, but it would be a start. It would take us a considerable distance past just muddling through.

In one of its primary aspects, disease must be construed as a biological event little modified by the particular context in which it occurs. As such it exists in animals, who presumably do not socially construct their ailments and negotiate attitudinal responses to sufferers, but who do experience pain and impairment of function … Nevertheless, it is fair to say that in our culture a disease does not exist as a social phenomenon until we agree that it does—until it is named.

(Rosenberg 1992, pages 179-180)

The initial response by many physicians was that the problems were not real. For a time they were dealing with a cluster of symptoms that had no name—and without a name there was, in essence, no disease.

(Halstead 1998, page 44)

One of the common tactics used for attacking the linguistic relativity hypothesis is to ridicule those who defend it by comparing them to people who believe in magical incantations. Whorfians, it is suggested, should really be called Shazamians, or Abracadabrians.

Let's set aside for the moment the mysterious cast that has been given to our world by the discoveries of quantum physics. We're aware now that the objects which surround us are actually made up of energy fields and busily moving particles, and are mostly empty space. But we don't perceive reality that way, not without the help of powerful electronic microscopes that are impossible to carry with us everywhere we go.

This whole semantic domain is one vast lexical gap in English; we have no vocabulary for talking about quantum reality in an everyday context. Our language reflects our ordinary perceptions; we refer to chairs and books as "concrete objects." As long as we restrict our discussion to such objects, the accusation of Whorfian leanings toward blatant superstition is in some ways accurate. I cannot think up some new concrete object that has never existed in my environment before—let's call it a "crawshbeck"—and cause one to appear by saying, "Shazam! Let there be crawshbecks!" That's very clear, and not a matter of dispute.

However, many things that are abstract have just as much power to create effects in human beings as concrete objects do. When a tree is in my path, I will step aside and go around it; that's one of the effects the tree has on me. I can't see or touch liberty as I can see and touch a tree, but liberty has dramatic effects on my behavior, nevertheless. I wouldn't hesitate to state openly my opinions about the Congress and the President, even if they were totally negative; that's one of the effects that liberty has on me. And creating abstractions out of next to nothing—abstractions whose fantasy status is then extremely difficult to prove—is a snap. English has an array of linguistic devices that can be used that way.

In the field of medicine this particular power of language—the power to use language to create abstractions with substantial effects in the real world—is magnified enormously, so much so that it comes very close to being magical. Consider, for instance the striking example which revolves around the phenomenon of menopause.

Case Study: The Menopause Transformation

Menopause is a normal stage in every woman's life if she lives past fifty or so. It may in many ways be a nuisance and an embarrassment and an inconvenience. But no one would feel obliged to see a doctor just because they were going through a normal stage of life, however annoying they might find it. Think of young human males; they go through a stage when their voices are "changing" and they often find that extremely annoying and embarrassing and inconvenient. But no one expects them to go to a doctor and request treatment for their condition.

Women don't "have menopause" or "suffer from menopause"; they go through it. Women "have arthritis" and "suffer from arthritis"; they don't "go through arthritis." But look at the following dialogues:

1. WOMAN: Is there something wrong with me, Doctor?
   DOCTOR: No. You're going through menopause.
   WOMAN: Well, thank goodness for that! I was afraid I was sick.
2. WOMAN: Is there something wrong with me, Doctor?
   DOCTOR: Yes. You're suffering from hypoestrogenemia.
   WOMAN: I am? What does that mean?
   DOCTOR: It means that you have an estrogen deficiency. But don't worry; we have very effective treatments for it now. I'll write you a prescription, and if you'll stop by the nurse's desk on your way out she'll give you some information that you can read that will answer all your questions. And then I'll need to see you again in a few weeks to check on how you're doing with the medication, whether we need to make any changes in the dosage … that sort of thing.

Calling menopause "hypoestrogenemia" and/or "estrogen deficiency" is nothing but talk. Naming, that's all it is. But any literate adult speaker of English who hears the word "hypoestrogenemia" knows immediately that it refers to a disease or disorder, and understands that "estrogen deficiency" definitely means "something is wrong with you." SHAZAM! ABRACADABRA!

This is a three-step process for skilled medical professionals. They begin with the ordinary "layperson" name for something ("menopause." "change of life"); they assign an alternative lofty-sounding name that bears no obvious relationship to the ordinary name ("hypoestrogenemia"); and they reserve one or more slightly less obscure additional names ("estrogen deficiency") for use if and when the layperson asks, "What does that mean?"

Telling a woman that she's "going through menopause" doesn't do anything remarkable; if it has any effect, it's probably one of reassurance. Telling her that she "is suffering from hypoestrogenemia" or "is hypoestrogenemic" and "has an estrogen deficiency," on the other hand, activates an entire cultural construct with effects that will dramatically change her personal reality. If, as often happens, she is approached with language like that used in typical articles on the subject, her image of herself as a fit and healthy human being is in danger. The April 1987 issue of The Female Patient had a piece titled "The Climacteric, Part I: Conquering Osteoporosis, Vasomotor Instability, and Urogenital Deterioration" (Fayez et al. 1987, page 62). The woman portrayed in this article isn't just moving on to another normal stage of her life—she is unstable and deteriorating and deficient, characteristics that aren't associated with fit and healthy people in our culture. She has a new role in life, created by the Menopause Transformation: she is now A Patient, and in medical terminology she is eligible to assume the "sick role."

Things that were nothing but unremarkable characteristics of her daily and nightly life are now symptoms. Patients, especially those with symptoms, are expected to seek treatment, even when there is no cure. Treatment means setting aside time she would have spent in other ways and spending it in doctors' offices, and in going to and from doctors' offices. It means taking medicines and/or undergoing medical procedures; it means setting aside money to pay for doctors' visits and medicines and procedures, money that would otherwise have been saved or spent for other things. People have a different attitude toward a woman who is a patient and "being treated" and "on medication" than they have toward women who don't have those characteristics. The sick role, in our culture, brings with it an elaborate set of characteristics that stretch far beyond the individual who assumes it; the whole family has its life changed and rearranged when one member takes on the sick role.

Creating this disease of hypoestrogenemia, whether from scratch or by transforming menopause into a disease, is close enough to magic to inspire respect in anyone watching the process. The woman in question is precisely the same woman she was before the incantation was pronounced over her by the doctor; her situation is precisely the same as that of a woman who is told only that she's going through menopause. But her life is not the same, nor is the life of her family and intimate circle the same. It's no accident that the register of English spoken by most physicians is called "MDeitySpeak." The doctor says, "Let there be hypoestrogenemia!" and there is hypoestrogenemia.

And does the doctor see that it is good?

The answer must surely be yes, because the phenomenon has moved on into another phase of development. Women are now being told to get ready to deal with yet another disease. This one is called "perimenopause" and is identified as the seven years or so immediately prior to menopause itself; presumably the term "prehypoestrogenemia" is waiting in the wings in case "perimenopause" alone lacks sufficient impact.

In the April 1994 issue of Ladies' Home Journal, Dr. Philip Sarrel of Yale Medical School reported that two-thirds of women working outside the home suffer "a moderate to severe effect on their capacity to function at work" because they have perimenopause (Laurence 1994, page 73). The article advises women that although no consensus on treatment exists as yet, they must begin "monitoring their symptoms." And according to the 1993 Clinical Proceedings of the Association of Reproductive Health Professionals, there were 21 million women on the verge of coming down with perimenopause in this country at the time of publication. That's a lot of patients, and a lot of patients' families. That's a lot of additional money that will go into the total spent for health care—a total we claim to be working desperately to make smaller.

The next stage is predictable; once women are accustomed to being treated for perimenopause, they will begin reading that a new "health problem" called "postmenopause" has been identified, and it will be suggested that they begin monitoring its symptoms. The phrase "postmenopausal women" is already common in medical and health care literature.

So far, no one has performed these MDeitySpeak ceremonies over middle-aged and older men, although both "hypotestosteronemia" and "perihypotestosteronemia" are available for "discovery." We have no way of knowing whether it would be as easy to convince millions of American men to accept the sick role for those two many-syllabled conditions as it has been to convince women that being "menopausal" requires a doctor's services. That has yet to be tested.

Steven Pinker, you will recall, claims that people aren't so brainwashed that you can pull off this sort of thing. All that's necessary in such cases is to "point out" to them what's going on, whereupon the alleged reality-shaping power of language will be shown up for the sham that it truly is.

Let's be very careful here. Women in America don't spend years involved with our cumbersome medical and health care system because they are unaware that "hypoestrogenemia" and "estrogen deficiency" are MDeitySpeak euphemisms for "menopause." They know that perfectly well. It's not some secret that's being kept from them by a wicked medical conspiracy. They're neither stupid nor ignorant. But the power of MDeitySpeak is so great that this knowledge does not keep one woman in a hundred—provided she has access to medical care because she's rich enough or insured enough or poor enough—from becoming A Patient. Nor does it keep her family members and associates from rearranging their lives so that time and money and transportation are available for her visits to the doctor. It doesn't keep menopausal women who lack access to medical care from feeling deprived and neglected. I know this from personal experience, and it's a source of perennial amazement to me. When other "menopausal women" learn that I've been menopausing without medical assistance of any kind for at least a dozen years, they are shocked. I've tried explaining to them why I behave as I do in this matter, and I am here to tell you that the act of Pointing Out The Euphemisms appears to have no magic in it at all.

Meanwhile, there are languages and cultures which have no idea that this stage of a woman's life ought to bring their medical resources into play. My daughter and I have discussed this topic with her husband (Lao/English multilingual) in much detail. Bounlieng assures us, unequivocally, that Lao has no word for menopause and that Lao women don't mark off this period of their lives in any fashion, either linguistically or culturally. "They're able to stop having babies without any help from a doctor," he says, and I'm sure that's true. You then have to wonder how a woman with native fluency in both Lao and English deals with this conflict between the two languages. Which one does she rely on?

When I've raised this question with the American medical professionals in my seminars and consulting practice, they've had no trouble answering it. As they perceive it, nobody—once exposed to the superior medical knowledge we have in the United States—clings to his or her previous beliefs. The assumption is once again that the outcome can be taken for granted, and that the immigrant who learns English and is thus "assimilated" will always say something along the lines of "Before I spoke English I thought such-and-such a thing; now I know better."

Let's be frank here, because no purpose will be served by tiptoeing around the truth: The position taken, and not just by doctors, is that people whose medical concepts are different from those of "Western" medicine are in most cases simply ignorant, through no fault of their own, and that the remedy is to provide them with information about how "real" contemporary medicine is done—at which point they will be grateful for being brought up to date at last, and will immediately abandon their previous ideas on the subject. Notice how the physician-authors in this quotation word their question:

In what ways does acculturation change the beliefs of patients of various ethnicities, i.e., how are the cultures of immigrants transformed and combined with the culture of their adopted country?

(Blackball et al. 1995, page 823)

The question posed isn't "Does acculturation change the beliefs … " and "Are the cultures of immigrants transformed … "; those results are presupposed. No room is left here for an alternative resolution in which patients simply decide that the American English version of events is nonsense and rely on the perceptions shaped by their native languages instead.

My own experience with native speakers of other languages—not only those who are patients but also those who are working as medical professionals in this country—does not bear out that confident presupposition. The error has serious implications for the health care system in a multicultural nation like the United States

For example, evidence for the placebo effect has made it more than clear that a substantial part of "being well" and "getting well" depends on the patient's faith in the doctor and the medical system. That faith also determines in many ways the degree to which the patient is "compliant" (that is, follows the doctor's orders); this is an equally substantial factor in the ultimate outcome for the patient. How well is medical care going to work for multilinguals when one of their native languages slaps an illness label onto some chunk of their experience that has no such label in another of their languages?

We know that some multilingual populations in this country decide either for their native medicine or for the American version and reject the other alternative. We also know that English/Chinese multilinguals living in the United States, by contrast, typically go to at least two medical professionals—one who is practicing Western medicine and another who is practicing Oriental medicine—as if to cover all their bases. Most native Navajos follow that pattern, going to both a Western doctor and a Navajo medical practitioner. (It does strike the Navajo as bizarre that it is the other system that is referred to as "Western," yes.) Presumably this doubling makes Navajo and Chinese-American patients' health problems even more expensive and time-consuming. What other effects does it have on their behavior, and on their health?

In their book Philosophy in the Flesh, Lakoff and Johnson offer an example that is relevant here. They note that various metaphors for numbers exist—NUMBERS ARE POINTS ON A LINE, for example, and NUMBERS ARE SETS. And they say that what people do is choose their metaphor for numbers on the basis of the kind of math they want to do at the time. Perhaps that's what multilinguals do when their native languages offer them conflicting metaphors; perhaps they choose on the basis of the kind of thing they want to do at the time. What would that mean for multilingual doctors and nurses and medical technicians, if that's what they do when they're the ones experiencing the conflict between two or more of their languages? We don't know; no systematic, controlled research on this issue has yet been done.

Christopher Lawrence says, "The focus of inquiry should be on how communities come to frame or see diseases … What is required is an explanation of how and why perceptions are structured as they are and how and why they change"(Lawrence 1992, page 53). I agree with him. How? And why? And how do we find out?

In my experience, conscious awareness of the way a language shapes perceptions usually occurs when its speakers are able to step outside their language in ways that come along rarely in everyday life. Remember the discussion in Chapter 2 of differing perceptions about horseback riding? Each time I taught that material in my classes, the same thing happened. The English-speaking students, after hearing how Navajo gives a surface shape to those perceptions, would say, "I never realized before that riding a horse means doing something to the horse!" The Navajo-speaking students, hearing how English handles the matter, would tell me that they had never before realized that the Navajo equivalent meant doing something together with the horse. It's a cliché to say that you can't see the forest for the trees, but where native languages are concerned it describes the situation with dead-on accuracy.

Native English speakers who are medical professionals are only likely to discover the presuppositions that underlie their own concepts and beliefs about medicine if they are exposed to those that are present in other languages. No mechanism exists in medical training in the United States to make that happen, presumably because it's not considered important. I have been told repeatedly by medical educators that "there's no room for linguistics in the medical curriculum."

Medical Models and Metaphors

Metaphorical thought is commonplace and inescapable; in itself, it is neither good nor bad … The use of a metaphor, however, becomes pernicious when it hides realities in a harmful way.

(George Lakoff, in Lakoff 1991, page 1)

Human languages around the world present very different models of what health and illness/injury/disability and medical care are, depending on the metaphors the language uses for these items and all their related phenomena. The range in medical models encoded in human languages around the world stretches from "All health problems are caused by evil spirits and bad thoughts" through "All health problems are punishments for sin" to "All health problems are caused by germs" and far beyond.

In his book The Psychobiology of Mind-Body Healing, Dr. Ernest Lawrence Rossi insists that all health problems are caused by communication breakdown. Since physiological phenoma happen in the body when the brain sends its neurotransmitters out with instructions, he makes a very strong case. I find it totally convincing, especially since the human brain is unable to distinguish between external and internal stimuli and will dutifully send the body "Prepare to flee or fight!" messages even for imaginary crises. If Rossi's definition were generally accepted in medicine, medical schools would have no choice but to find room for linguistics courses. But his metaphor is not the standard one in Western medicine, and "communications" courses for doctors continue to have little or no input from linguistic science. (I'm happy to say that my own medical seminars are exceptions, but it's a lonely business; the only area of medicine in which I feel that linguistics is beginning to be taken seriously is in emergency medicine, where the necessity to obtain accurate information in a matter of seconds has created in EM professionals a healthy respect for the science of language.)

The African-American culture of the Sea Islands of South Carolina divides nonfitness into two subgroups: "ailments," which are caused by natural forces or punishments from God and can be helped by doctors; and "conditions," which are caused by such things as devils and witchcraft, and can only be helped by spiritual advisers. (Hamadeh 1987). This, needless to say, is not the standard metaphor either. My Laotian son-in-law feels exactly the same way about the idea that "germs" cause illness as the average Western doctor does about "evil spirits" filling that role: He considers a belief in germs to be a foolish and unfortunate superstition. It's not that he doesn't believe germs exist—but he considers them to be the result of illness rather than its cause.

We're going to examine here just two of the most common conceptual systems of medicine around the world: one in which illness is understood as a state of being broken; and one in which illness is understood as a state of being out of balance or unharmonious. An example of the first system is American mainstream medicine; examples of the other would include Chinese and Navajo medicine.

The problems with definitions of terms, and with expressing terms and concepts across the languages, are severe. (This is a situation that you'll be familiar with after our struggles with putting the water babies concept into English.) I'll do my best to straighten things out as we go along, beginning with the choice of a cover term we can use to express the unwieldy set of meanings-and-perceptions covered by such words as "illness," "disease," "disorder," "injury," "disability," "sickness," and more. It's common to see illness defined as "an absence of health" and health defined as "an absence of illness." Going around that circle again won't help us. I'm going to use "non-fitness" as a cover term for all of those deviations from the hypothetical human state of perfect health and perfect well-being. Not because it's an especially elegant choice, but because it seems to me to be the broadest and most neutral one I could choose, much like using the term "variety" to express a neutral attitude toward a way of speaking a language. A sprained ankle is nonfitness; lung cancer is nonfitness; deconditioning is nonfitness; schizophrenia is nonfitness; pregnancy, in the American system, is nonfitness; crise de foie ("liver crisis"), in the French system, is nonfitness; susto ("loss of the soul"), in the Ecuadoran system, is nonfitness; and so on.

In 1991 Harriet Beinfield and Efrem Korngold published Between Heaven and Earth: A Guide to Chinese Medicine. Unlike a number of previous books I've read on the subject, this one does an excellent job of laying out the Chinese medical and health care model in a way that's both understandable and interesting for the Western reader. There's a striking difference, however, between the ease with which the American medical system is explained and Beinfield and Korngold's struggles to express the terms and concepts of the Chinese system in English. Here's a typical example, from page 17:

Western medicine is the study of how the human machine works. When people are like machines, doctors become like mechanics … In this schema, the body is reduced to structural parts, proceeding from organs to tissues, tissues to cells, cells to molecules. The doctor as mechanic separates the whole into parts in order to discern the nature, proportion, and function of each constituent.

Western doctors, the authors go on to explain, identify the broken or malfunctioning part, figure out what's wrong with it, and determine the cause of the problem. Successful treatment then consists of fixing that part, removing it, or replacing it. In the same way that the cause of a punctured tire can be identified as a nail picked up from the road, and the tire is either patched or replaced, the cause of a "broken" tonsil can be identified as a particular bacterium or virus, and the fixing is done by giving the patient a medicine—or, if the damage is too severe, by removing the tonsil surgically. If at some point Western doctors became convinced that the body has to have tonsils to function properly, the next step would be tonsil transplants or artificial tonsils.

This is very straightforward; it gives us a solid foundation for understanding how Western doctors proceed in their efforts to do something about states of nonfitness in their patients. It will annoy doctors who claim to practice holistic medicine, and rightly so; it will offend doctors who consider the comparison with so low-status an occupation as automobile repair demeaning. (Never mind the obvious fact that our lives are in the hands of auto mechanics just as much as they are in the hands of doctors.) Despite the justifiable objections, it accurately and clearly expresses the perceptions most Western patients have of their doctors and medical care.

Now let's compare this clarity with a typical quotation or two—in English—from the corresponding section on Chinese medicine:

In Chinese medicine, health is the ability of an organism to respond appropriately to a wide variety of challenges in a way that insures maintaining equilibrium and integrity. Disease represents a failure to adapt to challenge, a disruption of the overall equilibrium, and a rent in the fabric of the organism.

(Beinfield and Korngold 1991, page 36)

If the cause of disease is understood as imbalance, then the goal of treatment is to recover balance

(ibid., page 44)

These statements may be accurate, but they in no way answer the basic questions readers would have about what a Chinese doctor or health care worker does.

I turned to another excellent book on Chinese medicine, Ted J. Kaptchuk's The Web That Has No Weaver, for help, and found the same sort of thing:

All relevant information, including the symptom as well as the patient's other general characteristics, is gathered and woven together until it forms what Chinese medicine calls "a pattern of disharmony." … The total configurations, the patterns of disharmony, provide the framework for treatment. The therapy then attempts to bring the configuration into balance, to restore harmony to the individual.

(Kaptchuk 1983, page 4)

Although all these authors try hard, over and over again, to sum up this "harmonizing" process for speakers of English, they get nowhere. Until, that is, Beinfield and Korngold change their tactics. After a considerable amount of this floundering about, they give up on constructing careful, rational, "expository" sentences and paragraphs and turn to the much more powerful mechanism of metaphor. They had already established the English metaphor as A DOCTOR IS A MECHANIC. The Chinese metaphor, they tell us, is: A DOCTOR IS A GARDENER. And suddenly the lights go on! Because although we may have no idea what someone we could only refer to as a "harmony-restorer" does, we know what gardeners do, in precisely the same way that we know what mechanics do.

Richard Selzer (writer and surgeon) has provided us with what I consider to be the finest possible description, in English, of Asian medicine in action. In his essay "The Art of Surgery" he describes a grand rounds at his hospital featuring Tibetan healer Yeshi Dhonden, personal physician to the Dalai Lama. The task was to examine a patient chosen for him in advance by Selzer's hospital, a patient whose diagnosis was unknown to anyone except the doctor who chose her, and afterward to discuss the diagnosis with the Western doctors. Selzer describes Dhonden's taking of the patient's pulse, which was a very different process than the one we Westerners are accustomed to. It lasted thirty minutes, and when it was over Selzer writes of his new awareness that "I, who have palpated a hundred thousand pulses, have not felt a single one." Dhonden also examines a urine specimen provided by the patient. And that is all he does. In the discussion with the doctors afterward, he presents his diagnosis (through an interpreter) as follows:

Between the chambers of her heart, long, long before she was born, a wind had come and blown open a deep gate that must never be opened. Through it charge the full waters of her river, as the mountain stream cascades in the springtime, battering, knocking loose the land, and flooding her breath.

The doctor who selected the patient agrees; that is the diagnosis. But he states it in English: "Congenital heart disease. Interventricular septal defect, with resultant heart failure" (Selzer 1976, page 78). The two shapes of the diagnosis represent two metaphors; they express the same facts.

Metaphors are the most powerful linguistic tool available to us, and this is true for every language. The woeful lack of attention given to such skills as how to find and how to use metaphors is a serious deficit in our educational system. Metaphors can fill lexical gaps, as in the example we've just considered; they are one of the few options available for working to fill the lexical gaps that go with language about the inexpressible and unspeakable, as in attempts to describe mystical experiences. And far from being "only figures of speech," metaphors can do something that no amount of logical argument, however solidly bolstered by statistics and facts, can accomplish: Metaphors have the power to change attitudes—swiftly, without fuss, and often permanently.

If you are persistent and determined, you may be able to hammer away at people with lengthy logical arguments and statistics until they give up and say, "Okay, you're right." You may be able to make them do that more quickly by administering rewards and punishments, if you have that sort of power. But all it will get you, 99 percent of the time, is shallow surface agreement; often it's nothing more than a pretense of agreement, which they're willing to offer to get you to shut up. Underneath the surface, their old attitudes remain unchanged. A successful metaphor, by contrast, simply replaces and/or readjusts whatever was in their minds before; their attitudes shift immediately, and will remain stable until and unless an even more powerful metaphor comes along and overthrows the previous one.

The lifeblood of a Chinese company is guanxi [connections]. Penetrating its layers is like peeling an onion. First come connections between people and ancestors, then those between people from the same village, between members of the family, and finally between the family and close associates who can be trusted … Typically, the Chinese … are less concerned with what is written in the contract than in the actions that people take to meet their obligations as they emphasize guanxi.

(Cannon 1994, pages 332-333)

Suppose that you and I are involved in a business transaction. I want to sell you ten widgets, or ten hours of a service, and you're willing to buy. When the business in question is business as represented by Standard American Mainstream English (SAME), what that means in theory is that the following sequence will take place:

1. I describe the widget or the service to you, accurately and honestly.
2. I agree to a price for the widget or service—a price that covers my cost of business (including fair treatment for any workers and subcontractors and suppliers who may be involved) while at the same time yielding a fair profit.
3. I agree to a delivery time for the widget or service, with fair provisions made for such things as returns and refunds if the widget is damaged or the service is unsatisfactory.
4. I agree to a date by which you must pay my bill for the widget or service.
5. You agree to all of the items above.
6. We both follow through on our agreements, fully and honestly.

That's the theory. The practice deviates from that ideal so wildly that the theory can more accurately be called a fantasy. In practice, with rare exceptions, those on both sides of the transaction will have maximum profit as their goal and will do everything they can conceivably get away with to fudge on the terms of the agreement while still insisting that they're fulfilling it. That's how business (except Internet business—a topic we'll take up later in this chapter) is carried on in the United States today. That is "how the game is played," and profit is "how you keep score."

Don't be misled by the word "game"; we're not talking about "Ring Around the Rosy" here. Americans take games very seriously indeed. To bring that point forcefully home, you need only compare the annual salaries paid to our professional athletes with those paid to our professionals in education, law enforcement, and health care. Football, in particular, can be looked upon as almost a religion in the United States.

The core metaphor for business in the United States today is a curious meld of the two metaphors BUSINESS IS FOOTBALL and BUSINESS IS COMBAT; individual businesses will differ in how much they lean toward one or the other of the combined metaphors. What makes the combat also a game is the fact that business is a form of combat in which people not only aren't encouraged and rewarded for killing or injuring others, but are forbidden by law to do so. The American tobacco industry's disastrous problems in the late 1990s are a perfect example of what can happen when those in charge of a business lose track of that fact and begin following the specifications for a battlefield instead of those for a football field.

English is a superb medium for carrying out business as combat-with-limiting-rules. It's a superb medium for appearing to be following the rules while at the same time subverting them, the way a football player will pretend not to have the ball while at the same time running with it.

English (and especially SAME) offers its speakers and writers a varied assortment of mechanisms—linguistic tools—for this purpose. We're going to take a look at seven of the most common and productive ones. Not because this is a book about Standard American Mainstream English, but because SAME is so overwhelmingly the international language of business. There are signs that this may be starting to change at the very highest levels of global business, with some major firms beginning to demand multilingualism as a qualification for hiring top executives. But this trend is new, and it flies in the face of United States tradition; even if it becomes the norm, it will take a long time for it to develop and spread through our business culture. The standard situation today and for the immediate future continues to be one in which the SAME-speaking American businesspeople know only English, while the foreign businesspeople they're interacting with use English as a second language.

This is a source of serious problems. As you well know, the fact that two people from different cultures or subcultures are speaking the same words in no way guarantees that they're expressing the same meanings with those words. Even when the foreign speaker's English is extremely fluent, the appearance of a shared meaning can be an illusion; this becomes even more likely when the foreign English wasn't learned in the United States or from American English instructors. As Susanne Niemeier points out, the English used in such situations is influenced by the speakers' mother tongues in ways far more important than just pronunciation and "accent":

Their nonverbal behavior, for example, does not automatically switch to an "Englishized" nonverbal behavior … Thus, when they think the negotiation partner should have understood the (verbal and nonverbal) signs they are using, misunderstandings still occur because some signs may be differently encoded—and decoded—in the other's culture or may not be noticed to be signs at all.

(Niemeier et al. 1998, page 2)

There are now many Englishes, from all over the world, each one strongly influenced by the native language and culture from which it comes. They are not the same language. (For more details, see Chapter 7.) And this isn't just a "vocabulary problem," something that could be fixed if people would only learn more words and phrases.

The number of language-related potential problem areas for international business is so large that no one book (let alone one chapter in one book) could possibly explore them all; we'd need many shelves of books. The potential problem areas include at least these five, which we'll be discussing in this chapter:

• Different understandings of time and of space
• Different positions on what's right and what's wrong
• Different definitions for the key vocabularies of various semantic domains
• Different requirements for making information explicit
• Different dominant cultural metaphors

Why American English Is Such a Good Language for Business

Any semantic unit whatsoever can be expressed in any language whatsoever. We see this demonstrated whenever people say that a word or phrase "can't be translated into (some language)" and follow that up by explaining what the allegedly untranslatable word or phrase means. What they mean (as they would acknowledge if they were speaking more carefully) is that the word or phrase represents a lexical gap with no simple and convenient equivalent in the language, not that it actually can't be translated.

For example, there is a Navajo verb (ilhk'ddd) that can be translated as "taking a newborn animal that has been rejected by its mother because some human being was so careless and clumsy as to touch it with hands that had also and very recently touched some malodorous substance like bacon fat or grease or some such thing and putting that newborn animal to the teats of a different mother animal in the hope that the second animal will accept it." Certainly I can translate that into English; I've just proved that, although the translation is rough because I left out many details that would have exhausted the English-speaking reader's patience if I'd included them. However, if I wanted to translate into English a poem in Navajo whose first line was "At dawn I (some form of ilhk'ddd)," I would be facing a formidable stylistic challenge.

This linguistic comfort factor—what parts of reality a language makes it either easy or difficult for people to talk about—is one of the most significant ways in which languages differ. It's not likely that business documents and conversation are going to require much use either of forms of ilhk'ddd or their English translations. But there are similar, though less extreme, examples that are much more common. The Navajo names for things like car parts and hand tools tend to be so complicated and lengthy that many Navajo speakers feel obliged to use the English names instead, at least in business contexts. This is why, when listening to commercials for auto repair shops and car dealers on Navajo radio in the American Southwest, you hear a stream of Navajo punctuated by the English words "pliers" and "wrenches" and "mufflers" and "catalytic converters" and so on. It's not that Navajo doesn't allow speakers to discuss car parts with Navajo words; it does. But discussing them in a language which names such items as "that with which such-and-such-a-thing is done in such-and-such-a-fashion" takes a lot of time and patience. When you're doing radio commercials, which are charged for by the second, it won't be easy to work those Navajo terms in and still convince your advertisers that they're getting good value for their money. This isn't a matter of style or of translation skills; it's a matter of the bottom line.

It happens that English makes business discourse—that is, the language phenomena that go with buying and selling, and matters related to buying and selling—both easy and quick. To get through my explanation of why that's so, without tangling us up in a thicket of linguistics jargon and Academic Regalian, I need to introduce one term that may be unfamiliar to you: nominal. As follows:

A nominal is a noun (or any item that can do everything a noun can do) plus whatever language chunks go with it.

A typical English sentence will have a verb, plus one or more nominals that fill roles for the verb. For example: "The linguist recited nine boring grammar rules" has "recited" as its verb, plus two nominals, "the linguist" and "nine boring grammar rules."

"The linguist" fills the role of that-which-does-the-reciting; "nine boring grammar rules" fills the role of that-which-is-recited; the verb tells us what action is involved, that it took place in the past, and that it's over with. Every sentence with a verb, in any human language, can be thought of as potentially including a whole set of nominals, along these general lines:

Somebody does something, perhaps to someone or something, perhaps for someone, perhaps along with someone, at or during some time, in some location, in some manner, perhaps along some path in time or space, for some reason or purpose, perhaps using some instrument.

How many of those potential "somes" (and more) are actually spoken or written down, and how many appear not as indefinites but with explicit details, will depend on the rules of the language, the context, and the speaker's needs and intentions.

In scholarly literature on this subject the terms ordinarily used (taken from mathematics and logic) are "predicate" and "argument"; nominals are the arguments of the predicate. It might be possible to make these matters more difficult to talk about, but you'd have to work at it. This traditional terminology is much less efficient for ordinary communication than the Navajo names for car parts are, and I'll have nothing further to say about it.

With that established, we can go on to look at the set of characteristics that make English the businessperson's best current medium for business discourse, both spoken and written and—as in much electronic discourse—in between.

The English Advantage

Suppose you know that one of Metamega Oil Company's tankers spilled a substantial amount of oil in Lake Erie today. You could say so, with a batch of nominals specifying that the spilling was done quickly and carelessly and as a result of the skipper's bad judgment; you could name the skipper while you were at it. You could focus sharply on the guilty, saying, "It was Tom Smith, skipper of Metamega Oil Company's tanker Scrambled Star, who spilled oil in Lake Erie today." Many alternatives are available to you. But it's also perfectly acceptable for you to simply say, "Oil was spilled in Lake Erie today."

This last tactic puts the spotlight on the oil, as if it had created the mess all by itself. We do know that someone or something did the spilling, because that's presupposed by "oil was spilled," but English doesn't require us to make it clear; we can delete the spiller from the surface shape of the sentence completely. Similarly, we can say, "Mistakes were made" and "Flaws were overlooked," and "Funds were accidentally misappropriated," a strategy that Dorothy and Joseph Winters have named the "Passive Exonerative." (I am beholden to linguist Geoffrey S. Nathan, the Winters' son-in-law, for this elegant and apt label, which is used with the Winters' kind permission.)

This is extremely convenient, as demonstrated by the following classic example:

An accident has occurred at the Chernobyl nuclear power plant as one of the reactors was damaged. Measures are being undertaken to eliminate the consequences of this accident. Aid is being given to those affected. A government commission has been set up.

(The Soviet press agency Tass, quoted in the New York Times for April 29,1986)

This tactic isn't unique to English; all languages, it seems, allow it. Linguistic universals—that is, items that can reasonably be assumed to be present in every human language, such as vowels or nominals or a way of making things plural—are an important concern in linguistics. (Linguists propose them with great caution, because they are proposals about what qualifies as a human language.) I spent quite a while going to linguists all over the world for help in finding out how universal this phenomenon of "making the responsible nominals disappear" might be. So far as I have been able to determine, there is no human language in which it is literally impossible to do this. No language exists in which speakers would be violating a grammar rule if they didn't openly identify the "spiller of the oil." Some languages do it with passive sequences such as "Oil was spilled"; some do it by constructing sentences that would translate as "Somebody spilled oil" or "One spilled oil" or "They-unidentified spilled oil"; many do it with sequences that translate literally as "Oil spilled itself." However, English is extremely tolerant of this sort of thing, and the "Oil was spilled" tactic is only one of a number of possibilities it offers for hiding or eliminating nominals.

2. English lets you easily and efficiently hide information away in presuppositions and stacks of presuppositions.

Suppose my company has just put a new brand of laundry soap on the market. Chances are, it's much like all the other laundry soaps already available. If I want to say that our new soap is "technology-enhanced" (whatever that might mean), I'm claiming that that's so, and I'm likely to be asked for evidence. However, in English things that are owned are presupposed to exist. A sentence like "My black leather briefcase with the fancy zipper does not exist" would strike a native speaker of English as very strange, and could be used only in special contexts. (We can't say it could never be used, because you can always put together a sentence like "[X] is not a grammatical sequence of English.") This means that I can use a possessive word or phrase to presuppose that the soap is technology-enhanced and then finish my sentence with some claim too trivial to inspire argument. Like this:

My company's new technology-enhanced soap is now available in stores all over the country.

Or

Our new technology-enhanced soap is now on the shelves in stores from coast to coast.

I don't have to stop there. I can stack the presuppositions up as high as I feel a listener or reader will let me. Each presupposed item will turn what would have been an open claim on my part into information that I can reasonably expect you—linguistically speaking—to take for granted. Like this:

Our new and innovative scientifically designed technology-enhanced soap is now available in stores all over the country.

This doesn't guarantee that no listener will challenge me with a demand for an explanation of "technology-enhanced" or "scientifically designed"; it might happen. But it's not nearly as likely as it would be if I just made my claims straight out, especially if my sentence is in the middle of a speech or other presentation. When my listener is someone from a culture that disapproves of such challenges, they're even more unlikely; if the other person is reading my words rather than hearing them, and I'm not present, no convenient opportunity for a challenge exists.

3. English doesn't require you to specify the source of the information you put in your sentences.

There are languages—Aymara is one—in which every sentence has to include a chunk stating what it is that makes you feel justified in saying what you're saying or writing what you're writing. These languages have sets of forms (often called "evidentials") for specify ing the source of information. Evidentials include such things as having perceived something with your own senses, having learned it from a source you consider trustworthy, having made it up yourself from scratch, having seen it in a dream, and so on; which members of the potential complete set occur will vary from language to language. In the same way that English sentences are ungrammatical without some chunk that tells you whether the action or state was or was not in the past, Aymara sentences are ungrammatical without an evidential either explicitly present or made clear by surrounding sentences. Aymara speakers could of course be lying about their sources of evidence, but they have to do so right up front. They can't lie in the far more convenient English fashion.

Imagine what would happen in American business if statements like "This is the best insurance policy on the market" or "We know you're going to be glad that you bought this automobile" were ungrammatical without an evidential. Imagine what it would be like if the absence of an evidential in an English sentence was automatically taken as evidence that you were trying to deceive or mislead. It wouldn't make business discourse impossible, but it would make it vastly more complicated, and would require radical changes in the way we do business in America.

4. English has an enormous vocabulary—hundreds of thousands of words—with an abundance of items that are either opaque or so nearly meaningless that they can be made to mean whatever the user wants them to mean. (Like the word "interesting.")

In some languages you can look at most words and figure out their meanings from the meanings of their morphemes—their meaningful parts. This is much rarer in English, where the jargons and registers of fields and subfields are often meaningless even to other native speakers of English. If you choose a word or phrase of English and then realize that it's uncomfortably narrow and precise, so that it might box you into a business commitment before you're ready, you can almost always find a number of other lexical items to use instead. As Gerald Parshall says, "What makes English mammoth and unique is its great sea of synonyms, words with roughly the same meaning but different connotations, different levels of formality and different effects on the ear." (Parshall 1995, page 48)

It's easy in English to talk at great length, and in a way that sounds serious and significant, while saying almost nothing; this is one of the basic skills learned by everyone who gets a Ph.D. from an American university. Suppose I am presenting a seminar, and the participants are giving me a hard time by refusing to ask questions when I offer them opportunity. They may think that this will make the seminar shorter or force me to stop talking about my topic and start telling jokes to fill the time. They are wrong. I have a Ph.D. from the University of California-San Diego's linguistics department; I am highly trained in the technique of saying things like this:

Today's topic, when subjected to primary effects of interrelated and synergistic demographic parameters, offers us greater scope than would be provided by mere humanistic rubrics in a similar context.

So far as I know, that sentence is for all practical purposes meaningless, but I can use the lexical abundance of English to go on like that for hours. Many of the world's languages offer their speakers no such luxury.

5. English has an extraordinary willingness to add to its vocabulary, not only by coining new lexical items but by borrowing them from any and all other languages. In addition, it expands its vocabulary by using nouns as verbs, and verbs and adjectives as nouns, at the drop of a hat.

We've already seen that a willingness to add words from other languages is by no means a universal characteristic; many nations look upon borrowed words and phrases as a contamination of the native tongue. English has no such barrier, despite the frantic and dogged efforts of purists.

6. English has a writing system that contains almost no information about the tune the words are set to, so that a written agreement can easily be stripped of every last bit of the emotional information that would be present in a verbal agreement.

Most of the emotional information is carried in English not by the words that are said but by the intonation (including tone of voice) that creates the tune the words are set to. As a result, speakers of English can easily say a single utterance with many different emotional meanings, including one that contradicts its "dictionary meaning." A standard exercise in beginning drama courses is to give the students some random word like "goldfish" and instruct them to say that word in such a way that it conveys rage, then lust, then terror, then tenderness, and so on down a long list of emotions. English allows that all to be done by intonation, while many other languages would have to add words or pieces of words—with specific emotional meanings—to accomplish the same thing.

This has the obvious drawback of letting verbally abusive people get away with their unpleasant language behavior by saying, "But all I said was … " and following that up with the words they did in fact say, but set to an entirely different tune. William Labov and David Fanshel are all for this feature of English:

In our view, the lack of clarity or discreteness in the intonational signals is not an unfortunate limitation of this channel, but an essential and important part of it. Speakers need a form of communication that is deniable. It is advantageous for them to express hostility, challenge the competence of others, or express friendliness and affection in a way that can be denied if they are explicitly held to account for it.

(Labov and Fanshel 1977, page 46)

Well, yes, it certainly is advantageous, provided you are "winning" in the business transaction. And that is how the game is played.

It may be that everyone who was present in a meeting to negotiate a contract has understood that members of the negotiation teams were furious or suspicious, or were agreeing only with great reluctance. Nothing in the English writing system requires information of that kind to appear in the contract or other document that comes out of the session. This is a case in which it's true that "you had to be there"; otherwise, you may be entirely in the dark about the participants' feelings.

7. English appears largely free of special discourse restrictions that would make business discourse tricky and difficult.

Japanese has a constraint that goes something like this: "Do everything you possibly can linguistically to make sure that no one ever has to say no to a request." Japanese has built-in mechanisms for indicating the rank and status of everyone present in an interaction, mechanisms that can't be left out of the sentence. Chinese has a constraint which says that you must do everything possible linguistically not to make awkward developments explicit—because once they're explicit, something has to be done about them. English has no such inconvenient features; it no longer has even the distinction between a formal "you" and an intimate one, like the French vous and tu.

Kyoko Mori (quoted in the November 3, 1997, Publishers Weekly) says that she doesn't like to speak Japanese because "you have to agree on … which one of you is superior, how close you expect to be … and who defers." Respondent Atsushi Furuiye (Japanese/English) writes that "English is better than Japanese when negotiating MONEY … Japanese with its complicated ways to speak according to social seniority/inferiority is sometimes a pain." Respondent Fran Stallings (English/Japanese) explains that Japanese has "honorific syllables which can be inserted before and/or after a general word to elevate its status … And there are whole separate sets of vocabulary (nouns and verbs) to be used for different levels of status." Chinese has a number of different words meaning "no," each carrying with it various nuances of meaning about the intensity and finality—the true "no-ness"—of the "no" in question.

Speakers of languages that make heavy use of devices like these can feel linguistically naked when speaking English. They may react to that discomfort by trying to find English substitutes—thus often leading monolingual English speakers to perceive them as either absurdly and excessively polite or as faking politeness. This perception then colors the English monolinguals' judgments negatively and interferes with business.

The seven characteristics of English discussed above (and others not mentioned here) work together synergistically—so that the combined effect is greater than the sum of the parts. Perhaps the most useful aspect of that effect for business is to establish and maintain that quality of deniability we've been discussing. ("Who, ME? Me, carrying the BALL? What makes you think THAT?") English is supremely deniable. This is not the only reason it has become the obligatory language of world business (there is, for example, the extremely important fact that English uses a writing system easily handled by keyboards), but it certainly hasn't held it back.

As long as business transactions are carried on between consenting adults who are all native speakers of Standard American Mainstream English, the results of all this are acceptable. But what happens when other languages (including other Englishes) are involved? When non-native speakers of English are involved? When the communication takes place on the Internet, where English dominance is so extreme that the four other most common languages (German, Japanese, French, and Spanish) combined are only used about 10 percent of the time? Then what?

Some of the things that happen in these situations are amazing; they lend support to the idea that we human beings all speak dialects of Terran. Here is an actual telephone conversation (recorded by Alan Firth and reported in Newman 1995) between an Egyptian cheese importer and a Danish cheese exporter whose only language in common for business is English:

IMPORTER: So I told him not to send the cheese after the blowing in customs.
EXPORTER: I see, yes.
IMPORTER: So I don't know what we can do with the order now.
EXPORTER: I'm not, er, blowing, er, what, er, what is this, er, too big or what?
IMPORTER: No, the cheese is bad. It is, like, fermenting in the customs cool rooms.
EXPORTER: Ah! It's gone off.
IMPORTER: Yes, it's gone off.

What's striking here (and what fascinates linguist Firth) is the way meaning is negotiated on the spot for business purposes, while the two businessmen are talking. "First, the Dane acts as if he understands 'blowing.' Only when his sale seems at risk does he demand a definition. Then he provides a definition of his own, and the Egyptian acts as if he knew what 'gone off' meant all along."

English, the article tells us, "follows markets," resulting in a batch of special-purpose "Englishes" (more accurately, English registers) that we could call Oilglish and Cheeseglish and StockandBondglish and Hackerglish—and so on

This is an extremely complicated issue. Its parts are tangled up together, and they interact in complicated ways. In Chapter 7 we'll return to the topic of English as an international language. For now, in order to arrive at a rough understanding of the situation, let's briefly consider a few of the organizing factors.

The Language of Time

Business depends on time. We have a metaphor for that in English: TIME IS MONEY. The standard pattern for SAME-speakers specifies that when businesspeople agree to meet from 2:00 P.M. until 4:00 P.M., it means that none of them will arrive later than 2:05 without a really good excuse such as a major traffic jam, a flat tire, or an injured child. It means that everyone will feel obligated to stay until at least 4:00 P.M. unless a similarly good excuse for leaving is available or the group agrees that all the business that can be done that day has been completed. It means that there will be a list of tasks (the agenda) to be accomplished during the meeting, usually in a specific order; it means that one thing will be done at a time, and each one will be made as complete as is possible before anything else is started. Any changes in these constraints have to be negotiated, and English has standard scripts for getting that done. The concepts of arriving and leaving and beginning and finishing "on time," and keeping to the schedule so that everything is done "on time," have dearly understood meanings.

However, when someone from another culture or subculture agrees to do all these things and uses phrases such as "on time" to express that agreement, it's not safe to assume that a shared meaning for "on time" exists between you and that person. We've all heard talk about cultures and subcultures where "nothing ever gets done on time"—a description that has a strongly negative meaning for SAME speakers, with major effects on such real-world decisions as where investments are made, where factories and other workplaces are built, and which people are hired to do which work. SAME speakers perceive those about whom they can say that "nothing ever gets done on time" in a way that involves attaching all kinds of negative labels such as "lazy" and "unreliable," and worse.

This is an example of the power of language, not the power of fact. It's not ordinarily the case that people from the other culture care nothing about being "on time"; rather, "on time" has a different meaning for them. When doing business in Peru, according to respondent M. J. Hardman, you specify "whether something is hora peruana or hora inglesa. The latter will be closer to the specified time; the former is likely to be a half-hour to two hours later." The accepted importance of "making the most of your time" by doing only one thing at a time (including the idea that only one person should speak at a time), and by finishing every item on the agenda if possible, varies from culture to culture. As does the idea that everyone in a meeting will do their best to "save time" by moving everything along as quickly as possible. Charles Dubow, in an article about doing business in Russia, writes that Russians "regard business meetings as a cross between an interrogation and a tea party … Over the course of a meeting that can drag on for several hours, nothing may be accomplished" (Dubow 1998, page 143). Why? Because Russian businesspeople "will win by doing nothing longer than you are willing to wait." It's not safe to leap to the conclusion that the Russian behavior represents what we call "stalling" in English. It's simply an example of a business strategy that differs from the standard American form and reflects a difference in the way time is perceived and understood. Dubow is trying to explain behavior based on Russian-language presuppositions by writing about it in terms based on English-language presuppositions; the result, inevitably, is a distortion of meaning.

The best account of different perceptions about the meaning of time is still found in the excellent books written by Edward T. Hall, who coined the terms "monochrome time" (only one thing is done at a time) and "polychrome time" (two or more things are done at the same time). Hall writes of cultures whose time concepts differ so radically from those of SAME speakers that conducting joint business activities is almost impossible.

The monolingual mainstream American businessperson is likely to believe that a simple solution exists for this dilemma; I've heard it explained many times. All you do, he or she will tell you, is sit down with the others involved in the business interaction and explain what "on time" means in English and why it's necessary to understand "on time" that way, and then include in the terms of your agreement an acceptance of that version of "on time."

This is naive in the extreme. It relies on the idea that you can assume that your listeners will take what you say on such an occasion seriously. If you are that businessperson, you feel safe making that assumption because the others will see that you're "spending precious time" talking about always being "on time," and because your brief monologue on the subject is one of the tasks "on today's agenda" and therefore something to begin and carry out and finish, on time. But you're forgetting something: These concepts aren't parts of concrete physical reality like rocks and desks; rather, they are parts of your Standard American Mainstream English grammar, built into your worldview by the meanings of the words and phrases of SAME. The end result is all too often going to be that you will later claim that you were "wasting precious time" when you gave your talk and that the other people involved can't be trusted because they don't keep their agreements.

Hard as many find it to believe, from the point of view built into the others' native language three things have happened: You have explained that you expect everything to be done "on time"; they have agreed to do that; and that's what they're doing. Even if they consistently arrive hours (or days) after the scheduled starting hour, shift back and forth from one task to another in a way that seems to you to be random, and leave whenever they feel like it.

In The Trembling Mountain (Robert Klitzman's fascinating account of his investigation of the disease kuru among the Fore people of New Guinea), sentences such as "We agreed to leave at 7:00, but he didn't arrive until 9:30" and "We agreed to leave on Monday but no one turned up until Wednesday" come along every page or two. This difference in attitudes about time isn't caused by a lack of moral character; it's a problem of language. That makes it no less maddening to the SAME-speaking employer or supervisor!

There are cultures in which no project can begin until everyone is perceived as thinking the right thoughts. Hall describes that phenomenon in a Native American culture, and points out what it would mean for American business affairs if it were widespread: "We could no longer schedule everything in advance because no one would be able to tell how long it would take to have the 'right thoughts' " (Hall 1984, page 189).

One of the things I was sometimes called on to do as a college professor was to go to a classroom where courses in a Native American language were being taught and tell the instructors that the class had to leave the room. The instructors would tell me that they couldn't do that because they weren't yet through with the lesson. "There is still more to learn," they would say, and I'm sure that was true. I would explain that even so, there was another class waiting in the hall to come in and that the classroom time and space were now theirs. And the instructors would look at me and say, "Have they no manners?" The SAME perception of "manners" means making certain that everyone can follow a single agreed-upon schedule and make the most efficient use of time in that sense; the Native American instructors were equally concerned about good manners, but they defined them differently.

A primary goal of education for assimilation in the United States is that immigrants will do more than just understand, intellectually, the meaning of the vocabulary used in the Standard American Mainstream English semantic domain of Time. The goal is that they will also internalize that meaning and accept it as "the way things really are," along with TIME IS MONEY and MONEY IS HOW YOU KEEP SCORE. So that it won't be necessary to explain and negotiate every smallest detail of every transaction in terms of time, because they will be able to use the basic principles of SAME time to predict the guidelines for acceptable behavior in every business situation.

Serious and justifiable concern exists about the question of whether—if education is conducted in the native language or dialect rather than in English—that internalization can happen "in time" for the person to become a self-supporting member of U.S. society who contributes to the economy in the way that all good citizens are expected to contribute. This is the kind of concern that underlies the behavior of the immigrant parent who, hearing the kids speaking their native language, asks, "What's the matter with you? You want to end up spending your whole life on welfare?"

The introduction of the computer as an obligatory feature of American business is causing some drastic changes in the perception of time, with potentially dramatic consequences. We'll come back to that later in this chapter.

The Language of Information

The term "presupposition" is used in different ways in different fields; even within linguistics, people argue about its precise definition. In this book we're using it to refer to information that every native speaker of a language knows is part of the meaning of some unit of that language even when it's not explicit in the unit's surface shape. Like the fact that "John stopped reading the book" presupposes "John started reading the book." In business, presuppositions are straightforward bottom-line issues: For every negotiation or transaction, decisions must be made about what other people have to be told and what you can take it granted that they already know. Decisions such as the amount of resources you have to put into training employees, and the amount of resources you have to put into meetings and business-related social events, depend crucially on this factor. Everything that you can presuppose in business will save you precious resources of time, money, and energy.

Edward T. Hall talks about presuppositions in terms of context, dividing cultures and subcultures into those he calls "high-context," in which people routinely assume that other adults already know a great deal, and "low-context," in which people assume that others know far less and have to be told a great deal more. "To give people information they do not need is to 'talk down' to them; not to give them enough information is to mystify them" (Hall 1984, page 61). Germany is a low-context society; Mexico and Japan are high-context. SAME-speaking American society falls between these extremes, less low-context than Germany and less high-context than Mexico and Japan. Only in monolingual single-culture business interactions is it reasonable to assume that everyone involved will roughly agree on what is presupposed and what must be explained. Beyond that sheltered circumstance, mistakes can be extremely expensive and bad for business.

When I was working with Native American languages, I would sometimes go to my Indian consultants' homes for work sessions lasting for hours. It was thirsty work, but I was never offered anything to drink, and I stayed thirsty. It wasn't until both I and my consultants had spent months of thirsty hours together that I finally learned what was going on. They took it as presupposed that I didn't want anything to drink, although they found it strange. I was an adult, and adults know that when they're thirsty they're expected to go help themselves to whatever is available. To ask me, "Would you like something to drink?" would have been rude; it would have been treating me like a child. On the other hand, it would have been rude for them to get drinks for themselves when their guest wasn't thirsty—this is something else that "everyone knows." Matters were made worse by the politeness constraints of my own subculture—Ozark, and southern—which obliged me to say nothing, month after month, no matter how odd the whole thing seemed to me. In a social situation a communication breakdown of this kind is usually only annoying; in business, it can be catastrophic.

Suppose I meet with you and tell you about a new clause that has been added to the contract I have with you; suppose I end my remarks with "Is that all right?" and you say yes, in English. Do we then have an agreement? It depends. For me, English "yes" means that you agree. Responding to your "yes" with "Does yes mean that you'll do it?" would be not only an unnecessary utterance but a hostile one; I would be treating you as if you were a child or had some other impediment to doing adult business competently. But if in your culture it's rude to answer a direct question or a request with "no," your "yes" may mean "I'll think it over and get back to you with a decision." It may even mean "Absolutely not." And why am I not justified in insisting that you say exactly and explicitly what you mean? Because we're both adults and you expect to be able to trust me to understand things that are known to every adult! How can we do business together, otherwise?

It will be obvious that no decision you can make in international business will be as important as the way you word every token of language used to transmit and explain your decisions, ask your questions, and respond to the language of others; in speech, the body language you use to accompany your words will also be crucial, especially if you are speaking American English.

The Language of Right and Wrong

When you're doing business with people from a nation that's not yet industrialized, you will be prepared to run into major ethical and moral differences. The Fore people that Robert Klitzman hired and worked with came from a culture many of whose members considered a ritual form of cannibalism both moral and ethical. Because it was cannibalism that transmitted the disease he was investigating, this was a factor he couldn't ignore; he had to face it and take it into account daily. When you're dealing with people from an industrialized culture, however—even though their native language is not English—you not only feel confident that they disapprove of all forms of cannibalism, you're likely to assume that they share your understanding of terms such as "being fair," "telling lies," "cheating," "fulfilling obligations," "being in debt," "being loyal," "stealing others' property," and many more of the same kind. You feel no need to ask them if they think it's okay to cheat in a contract negotiation, just as you feel no need to ask them explicity whether they think cannibalism is a moral act, and for the same reasons.

Unfortunately, your assumptions (with the exception of assumptions about cannibalism) aren't safe, not even when every aspect of the business transaction takes place in English. Anthony Judge, writing on the Web site of the Union of International Associations, advises that where written agreements are concerned we need to divide cultures into three groups: those in which signed written agreements are automatically enforced; those in which such agreements may or may not be enforced, "according to circumstances and political convenience"; and those in which such agreements are looked upon only as public relations gestures, with intentions about enforcement not even being relevant (Judge 1994, page 4).

Even between subcultures within the United States that do have English as a shared native tongue, assumptions about shared presuppositions on key moral and ethical terminology aren't safe. When an additional language is involved, taking such shared meanings for granted is a genuinely foolish move. The question is of course what to do about this problem. Nobody wants to do business with people whose actions baffle you, whom you don't feel that you can trust, and whom you may have good reason to believe feel exactly the same way about you.

You can read stacks of books like Scott Seligman's Dealing with the Chinese and Haru Yamada's Different Games, Different Rules: Why Americans and Japanese Misunderstand Each Other. You can read masses of articles from the professional literature of international business and management and diplomacy. You can spend many hours asking native speakers of the other language friendly, probing questions. You can take seminars galore on doing business with "the French" or "the South Africans" or "the Europeans." You can spend weeks exploring the many sites on the Internet that are devoted to these matters. You can do all these things—and still go into an international business interaction ignorant of the half dozen items it's most critical for you to know about the moral and ethical positions of your opposite numbers. Why? Because human beings, no matter what language they speak, tend not to be consciously aware of such things, and information that's not in conscious awareness is very hard to get at.

You'll remember my account in Chapter 3 of the decades-long misunderstanding I had about people and their television sets. I assure you, if I had asked those people bluntly, "Why do you leave your TV set on while we're trying to talk?"—something my southern Ozark culture wouldn't allow me to do—there's almost no chance that they would have looked at me in surprise and said, "Why, we do it because our TV set is today's equivalent of the hearth fire in homes of the past!" They would have been embarrassed and baffled; they would have said something like, "We just always do leave it turned on. Don't you?" And things would probably have gone downhill from there. We don't have easy access to that sort of information about ourselves unless someone has made it explicit for us, as I've done in this book. If we did, there'd be no need for psychiatrists and psychotherapists.

My personal recommendation is that whenever you do business across two or more cultures or subcultures, you should always hunt for the one piece of information most likely to reward you with the answers you need: the dominant metaphor or metaphors that the people involved in your interaction rely on when doing business.

If you ask an honest American businessman whether he believes that it's all right to lie, he'll say no. If you ask him whether it's all right for a football player to pretend he has the ball when he doesn't have it, he'll say yes. If you ask him whether it's all right to inflate sales figures just a little to impress a potential client, he'll say yes. If you suggest that the second and third answers contradict his first answer, he'll tell you that "they're not the same thing."

A visiting anthropologist from Mars might try to figure this out by making long lists of acts that the businessman does or doesn't consider to be lying. Human businesspeople don't have time for that. Suppose that they know, however, that the man's dominant business metaphor, through which his perceptions in a business environment are filtered, is BUSINESS IS FOOTBALL. In a football game it's not a lie to pretend you have the ball when you don't have it, or to pretend that you're going to run in one direction when you're actually going to run the opposite way. It is lying, in a football game, to dress up in the uniform of the other team. Knowing that business for this American is like football will let you predict the meanings he will assign to all those key terms listed on page 121 *** and many more, without having to memorize each one individually.

In Understanding Global Cultures, Martin J. Gannon devotes sixteen pages to explaining that the metaphor of the family altar has the same status for the Chinese that the football metaphor has for mainstream Americans, and outlining the effects of that metaphor on business transactions. My Chinese questionnaire respondents don't all agree on the accuracy of this idea, just as some SAME speakers would quibble over the football game metaphor if asked directly. But it's a place to start and something to investigate, and it's overwhelmingly obvious that interpreting moral and ethical terms in the context of a football game or football season will be radically different from interpreting them in the context of a family altar. It's also obvious that when an American businessman's dominant operative metaphor doesn't seem to be football, it is far more likely to be basketball or chess than it is to be a family altar.

Gannon's suggested metaphor for Japan is the formal Japanese garden, but he cautions that this means not an absence of conflict but conflict managed according to very specific rules and with attention to maintaining a surface impression of calm and control. In an interview in Forbes ASAP (Karlgaard 1993, page 73), a Fujitsu executive is quoted as follows: "You Americans think there is room for everybody in the market. We in Japan think that our competitors are taking rice out of our children's mouths. We think that anything less than 100-percent market share is not enough. We believe in Genghis Khan's dictum: not sufficient to succeed—everyone else must fail." By contrast, when Karen Mitchell discusses Haru Yamada's Different Games, Different Rules (Mitchell 1998), she explains that although both English and Japanese use Aesop's fable about the ants and the grasshopper, they end it quite differently. In America the fable ends with "the ants scorning the foolhardy grasshopper"; in the Japanese version "the ants invite the grasshopper in to share their winter meal, as they appreciate how his singing spurred them on during their summer labors." It would be more useful to investigate how these two seemingly contradictory attitudes fit together, in the Japanese business domain structured by the garden metaphor, than to memorize a list of "tips and hints" for doing business with the Japanese.

Finally, as is true with perceptions of time, doing business successfully will be impossible if speakers of one language constantly leap to the conclusion that differences result from deficiencies in speakers of other languages. The first hypothesis should always be that the problem is not in the person but in the language, in which case language offers the only valid solution.

The Language of the Internet

Ten years ago this heading probably wouldn't even have appeared in a discussion of the power of language for business. Today it's obligatory. Selling automobiles and computers and videotapes on the Web may be much like selling them in a physical dealer's place of business, but selling what we now call "intellectual property" on-line is radically different. As Esther Dyson says:

We are entering a new economic environment—as different as the moon is from the earth—where a new set of physical rules will govern what intellectual property means, how opportunities are created from it, who prospers, and who loses. Chief among the new rules is that "content is free." While not all content will be free, the new economic dynamic will operate as if it were … The provider's vital task is to figure out what to charge for and what to give away.

(Dyson 1995, page 137)

Content is free? What to charge for and what to give away? (The prototype example to clarify this is the standard practice of giving away a computer software program free but charging customers a fee for updates and for each time they need help using the program.) This, together with the idea that score is kept in terms of hits (number of visits to your Internet location, whether any business is transacted during the visit or not), and that your company and its entire plant can be totally "virtual"—existing inside your computer, with no fixed physical location in either time or space—is causing cognitive headaches for businesspeople on a massive scale. The older the businessperson, by and large, the more frequent and severe the cognitive dissonance is. The limits to resources on the Net, Dyson says, are not limits on time and space, but limits on the amount and degree of human attention. "Does a place in cyberspace exist," she asks, "if no one visits it?"

Past attempts to establish an international language, much less an associated international culture, have always failed, but the Internet shows signs of doing what no amount of effort along those lines has ever been able to do before. The language of the Internet is truly an international language, and it looks like written English. But it's not the same English that we're used to, it's Netglish, and its presuppositions are different. Questions such as "Where is your company located?" and "How long will it take everyone to get to the meeting?" and "How much does your widget cost?" and "How large were your profits last year?"—questions that are the heart and soul of off-line business—are often irrelevant for this new variety of English.

There was a brief period when it looked as if Netglish might establish one feature that would put a small dent in the deniability of written English. This feature was the emoticon, a little graphic squiggle that people used in an effort to give a surface shape to meaning that in speech is carried by body language. For example:

"I'm smiling as I write this." :-)
"I'm frowning as I write this." :-(
"I'm very sad as I write this." :-C

The Japanese found the sideways orientation of the emoticons perplexing, and worked up their own versions. "I'm smiling as I write this," for the Japanese, is written as ^_^ and has a version which identifies the smiler as "a girl," written as ^.^ (Pollack 1996).

For a while, everyone was using emoticons, and increasingly more elaborate ones. Some people began introducing them into off-line writing because they now felt linguistically "naked" without them, like Japanese and Chinese speakers without their linguistic status markers. I thought the practice might leak into ordinary written English and be taken up by people who had nothing to do with computers; if that had happened, it might have made a dent in English deniability. But emoticons seem to have been a passing fancy, and they are already disappearing from Netglish.

When my oldest was first learning how to talk, he came toddling across the room and asked me for a cookie in Spanish. When I did not react, he got mildly upset. Then after a five-minute tantrum, he returned and asked me for a cookie again, but this time he used the English word. After that, he never spoke Spanish to me again.

(Pamela Faber; personal communication)

We know that a number of different positions on the benefits or dangers of multilingualism exist among Americans today, along with many variants of those positions; we know that at least in the United States they are based in large part on folklore and politics and inadequate information. In this chapter we will set aside the ongoing arguments about whether multilingualism should be desired, deplored, or ignored. Temporarily, let's just take multilingualism in the home as a given, the way we take furniture in the home as a given, and turn our attention to the power of language in that context.

Which Language, When, and with Whom?

In the United States in recent years, governments at the federal level and below have been busy trying to establish regulations about which language—usually English—is to be used in schools and government agencies. We've seen a few examples of large businesses (usually, but not always, in the medical field) setting up regulations forbidding employees to speak a language other than English. Similar efforts can be observed in other industrialized countries around the world. However, no government in any of the major industrialized countries seems to make any serious effort to regulate the choice of which language is to be used in people's own homes, in private. The choice appears to be left up to the individual family, which doesn't always find it easy to make.

But Which Language Do They Use at Home?"

There was a time when multilingual households in the United States were almost exclusively of three kinds, making the answer to this question obvious once the basic circumstances were known. There were, in those days:

• Impoverished immigrant households, where most family members spoke only the language of their homeland, with perhaps one person knowing a little English.

In this situation, unless the family happened to live in an area with a very large immigrant community of its own heritage (usually only in big cities), learning English was a matter of survival. Even in the large cities, public education meant that the kids would have to learn English quickly and well to function in school; no bilingual education program existed. In such households the native language would be used within the immediate family, with English reserved for outside the home.

• Very privileged and wealthy households, where everyone spoke English plus additional languages as a matter of course because this was considered to be an obligatory part of an adequate upper class education.

In this case the family could choose any arrangement for language use that suited its fancy. If that arrangement was to speak Latin at the breakfast table, French at lunch, and English at dinner, with afternoon tea or cocktail conversation in Mandarin Chinese, that was perfectly fine. It was a matter of individual choice. Such households were, and still are, rare.

• Native American households, where older members might speak only their Native American language, where children spoke English, and with wide variation in whether younger members also spoke the Indian language. (The native peoples of the United States are divided on the question of whether they prefer to be called Indians or Native Americans; the decision varies not only from tribe to tribe but within the tribes as well.)

This was of course a special situation, in which—although it might be that speaking the Native American language was the most critical linguistic imperative, because the language itself was in danger of dying—extreme pressure to use English in the home instead came from many sources.

These stereotypical situations continue to exist in the United States, but in today's global society they're no longer the only common possibilities. Now that the ranks of professions and careers are filled with people from every part of the world, many multilingual American households fit none of these three stereotypes. Similarly, all over the world there are households in which one spouse (or both) is a native speaker of American English and one spouse (or both) also knows an additional language or languages.

I was surprised (when I sat down and listed it for this book) at how much multilingualism exists within my own extended family, even though it contains no diplomatic or similarly "international" households. One son knows both French and English. One daughter is a native speaker of English who also knows French; her husband is a native speaker of Lao who also knows Thai and English. Another daughter is a native speaker of English who knows French and who (although her skills are now rusty from lack of practice) once knew American Sign Language well enough to teach university courses in that language; her husband is a native speaker of English who also knows French. One nephew is a native speaker of English whose wife has both English and Spanish as native languages; his father, although in his late fifties, has begun intensively studying Spanish because his future grandchildren will have Spanish as one of their languages. Another nephew is a native speaker of English; his wife's native language is Russian and she also knows English. In three of these households there are children, and children are planned in a fourth. My stepson is fluent in English and Spanish; he lives with his grandmother, who is also English/Spanish multilingual. When I am added to the list, with fluent English and French—plus several other languages (including Navajo) which I can, as David Crystal puts it, "make use of"—there's quite a linguistic array. Not one of the households involved falls into the three patterns listed above.

All multilingual households, stereotypical or not, must come to an agreement about two basic questions:

1. What language(s) will the spouses use with one another?
2. And—if there are children—what language(s) will each parent use with each child?

The answers to the "which language" questions occur in astonishing variety. There are of course families in which one spouse speaks only Language X, and no one else who speaks a language other than Language X lives in the home. In such cases, even if the other spouse is multilingual, the decision will be essentially automatic: Language X will almost always be the language used at home. Except in that specific situation, however, all bets are off. Contrary to what might be expected, it doesn't seem to be the general case that a single language is chosen and dubbed "our family language." Nor is it always the case that the father speaks Language X to the children in the home and the mother speaks Language Y to them, or that mother and father speak Language X to one another and Language Y to the children. Things turn out to be much more complicated than that.

On my questionnaires, the most common responses to "Which language or languages do you use at home?" were "Both of them" and "All of them." This doesn't mean that they use the "wealthy family pattern" and give each language some regular time slot in the day. Instead, it means, literally, that all the languages are used all the time. Often more than one language is used at the same time, creating a language environment that's likely to strike the monolingual person as almost a guarantee of a disorganized and chaotic home life. But that's not at all the way the families living in these households perceive the situation.

Four Characteristics of Multilingual Homes

When all the data are examined, four interacting characteristics of multilingual family life run consistently throughout my database of questionnaire responses, with very few exceptions. Looking at the set will help us understand how having two or more languages regularly in use in the same home at one time works, and why it need not result in chaos.

The first characteristic is the casualness of it all. No one in my database reports being forced to learn or use some language in the home against his or her will, or having forced anyone else to do so. No one reports any personal trauma or unpleasantness associated with multilingual communication at home, although some recall unpleasantness of one kind or another at school. No one appears to find his or her situation unusual, even when it is very far indeed from the monolingual "norm." Waruna Mahdi writes, "The language we spoke at home was English between brothers and sisters, English and some Dutch between us kids and our parents (who spoke Dutch with each other), and the respective native language (Indonesian, Thai, Chinese, Russian) with household personnel. This is part of normal embassy life; my father was a diplomat." The families themselves report no perception of their linguistic circumstances as chaotic or disorganized or dysfunctional, or even difficult.

The second common characteristic is that some one particular language is the one in which strong emotions, positive or negative or both, will usually be expressed in the home. One respondent (with a very reasonable request to remain anonymous) wrote, "Any sort of intimacy with my wife is always in Spanish. We simply wouldn't be able to make love in English, even though we often use English together for other matters"; he adds that he and his wife couldn't have a fight in English, either. Many of my respondents note that when they're angry or upset at home, or simply "emotional," they always fall back on one particular language.

The third characteristic, which may come as a surprise, is a frequent report that the people involved often aren't aware which language is being used unless something happens to bring that fact specifically to their attention. For example:

Margaret Ann Doty (English/German, living in Germany) writes: "With my son I only use English at home, unless he has German friends visiting or we are watching German TV. Sometimes we laugh when we realize we're speaking German. It can be an unconscious act. I watched a Tom and Jerry film we rented once and wondered why they were singing in German … At that moment I wasn't conscious of which language I was understanding, but if someone had asked me I would have sworn it was English."

Colin Whiteley writes, "It is quite possible to talk, read and write without necessarily being able to remember afterwards which language was used. Recently I received a two-page letter from my son, with one page in English and one in Spanish. I only realised this after reading it several times."

Whiteley goes on to make a claim that seems to me to be extremely important: "When you are concentrating on the real topic … the language is totally transparent, and only the meaning is 'visible.'" I've added the emphasis; I think it's well deserved.

I can add my own endorsement for this characteristic, and add an observation. When I was living with my first husband's family in Geneva, the language we used when the entire family was together was French. I enjoyed the French conversation and took part in it without hesitation. I quickly lost any conscious awareness that I or anyone else was speaking French. However, in the middle of such conversations it sometimes happened that someone would speak to me in English—a guest, for example, who realized from my abominable French that it couldn't possibly be my best language, and could tell from my body language that I was an American. When that happened, I would suddenly find myself unable to speak either language. Both French and English would suddenly turn into meaningless noises and it would take me a while to recover and be able to answer the well-intentioned guest.

The fourth (and to me the most interesting) characteristic is the association of each language with a particular person, and/or the association of each language with a particular role filled by that person. For example:

Respondent Wen-chao Li writes, "When I was in Hawaii, the pattern was that I spoke English to the kids at school, and Mandarin to the adults at home … My sister fell under the category of 'kids at school,' so to this day I always speak to my sister in English, and to my parents in Mandarin, even if we're sitting at the same dining table, taking part in the same conversation." Recent literature on multilingualism routinely mentions this characteristic, and my respondents' comments confirm the resports.

Pamela Faber's four children acquired three languages from birth (her English, her husband's Spanish, and French at their elementary school in Paris); she says she was "always amazed at how early they learned to assign languages to people. Roughly by the age of fourteen months, they had realized that their mother had one set of words and their father had another … They have associated languages irrevocably with people, not with contexts."

Peter Yongqi Gu writes, "I never speak English with my wife (also a Chinese with excellent command of English) unless there's a friend around who doesn't speak Chinese. I never speak Mandarin with my parents and folks at home; I speak Hebei dialect to show intimacy. When I realize, for instance, that the guy at the other end of the line is my brother, we both switch to Hebei dialect at once."

Suzanne K. Hilgendorf (English/German respondent) writes, "My first long-term relationship was with a German, and that language dominates that domain still."

Colin Whiteley (English/Spanish, and more) agrees, and writes, "I have often noticed, with myself and with others, that the language used by two people who share more than one language is always that of the first meeting; i.e., once the mode of speech is established, it is very unlikely ever to change."

James L. Fidelholtz (English/Spanish) explains that at home he speaks "English almost all the time with my wife, about 90 percent of the time with my two older children (seventeen and fourteen) and very occasionally with my youngest child (seven)" ; with that youngest child he ordinarily speaks Spanish. The different practice with different children is because after the family's return from a two-year sabbatical in the United States the older children insisted on speaking English with their parents.

Carol Troen (English/Hebrew) identifies English as the language used in her home—but there are exceptions. "With my older granddaughter," she writes, "I speak Hebrew and English; with the younger, Hebrew almost exclusively … My kids often switch into Hebrew when they talk among themselves … When I help our youngest with homework, we speak Hebrew." And she says, "I remember when Judah was born (child Number 5) a few months after we arrived in Israel. I was conscious that I had to speak English with him because I didn't know how to speak to a baby in Hebrew … The part of me that knew about children and knew how to talk to them could only speak English; gradually an act of 'translation' has occurred across these lines."

Pamela Swan notes a similar phenomenon; she would want her children to learn Japanese, a language that was always an "outside the home" language for her, but "I would like to tell my kids that I love them as I tuck them into bed, but it would be difficult for me to do that in Japanese."

Carol Troen also explains that "man-to-man" talk at home—"army talk," for example—among her adult sons and her husband is usually in Hebrew. I assume that if they were speaking with guests who were female soldiers, about army experiences, they would also speak Hebrew; nothing in Troen's account indicates that this choice is based on sexual gender. (However, a number of my male respondents mentioned that, when speaking at home in a situation where only other adult males are present, they prefer a particular language; the reason stated can be summarized roughly as "It's easier to make jokes in that language." )

Finally, here is an anecdote that is not only linguistically fascinating but also seems to me to sum up the subject perfectly:

My young son and I moved to Indonesia when he was two years old. By the time he was three, he was bilingual in English and Indonesian (spoken to him by the servants). He jumbled the languages for about five minutes. After that, although I could speak to him in Indonesian and he would understand me perfectly, he would reply to me only in English, and to the servants only in Indonesian: he had sorted out who spoke what to his own complete satisfaction. Of course, this meant he thought there were two languages in the world and he spoke both of them. I'll never forget the astonishment on his face when, at the self-assured age of three and a half, he looked at a Korean woman, categorized her, and spoke to her in Indonesian. When she replied in Korean, he was utterly dumbfounded."

(Kathryn Berck, personal communication)

When I discuss arrangements like these with people from traditional monolingual homes, the reaction is always one of astonishment that they don't lead to endless confusion in the home, as well as to friction and frustration. How, they want to know, could such arrangements possibly work? I can tell them only that, however strange it may seem, they do.

"But Doesn't All This Switching Around Confuse the Children in Multilingual Homes?"

No, fortunately it doesn't. If the children ever experience the state of knowing no language at all that I would drop into when languages shifted, they quickly get past that stage. The only hint of such a thing in my database comes from Michael Paris (English/Polish/Polish Sign Language), who writes that he has "gone through the experience of having my perceptions slide out from under me thanks to the language I was using." Both Paris and I began acquiring our other languages well past childhood; people immersed in two or more languages in infancy and very early childhood appear not to have such experiences.

When we change words we invite a changed perception of the reality to which the words point. To demand that the words of faith change is to demand that one's faith change. And in changing it can either grow or shrivel, blossom or die.

(Barr 1988, page 366)

Grammar and attitudes have nothing to do with each other. Benjamin Whorf thought they did, but he has been proven wrong.

(Beaver 1988, page 325)

We have considered in some detail the power that language has in medicine, in business, and in the home; we've discussed many aspects of its power in education. Now we're going to turn to the often very different power of religious language. The two quotations above make opposing claims. The first claims that the linguistic relativity hypothesis is true, while the second (from a linguist's letter in response to the first) declares it false. Knowing which of those two claims is valid is arguably more important in the context of religious language than in any other area of life. Religion and spirituality permeate every culture and every worldview, using religious language and talk about religious language as a primary mechanism of transmission. It's difficult to overestimate the power of such language over our lives, even for those who claim to have no religious beliefs and to be part of no religious tradition. In much of the Western world, religion is the area of life in which the most elaborate cognitive gymnastics are required, because of the seemingly unresolvable conflicts and contradictions.

What Is Religious Language?

Religious language can be defined in many ways, from casual conversation ("I believe in God, I guess, but I don't believe in angels") to sacred texts and purported revelations. The task of definition is made difficult by the fact that no method exists for determining whether a sequence of religious language is valid, that is, for distinguishing genuine religious language from phony religious language. When I gave lectures on this subject at the university, my students, whether they described themselves as religious or not, were startled to learn that this was so. They told me that they'd never thought about the matter before, but now that it had come up they felt that surely there must be some recognized standard against which alleged examples of religious language could be measured for validity, as well as some recognized international entity with the power to apply those standards and rule on such questions.

There are no such standards, and no International Board on the Validity of Religious Language exists. Let's suppose that I've walked into your living room and announced that I've just had a religious revelation, which I proceed to recite to you in the form of a creed or a psalm or a prophecy. You will probably think I'm out of my mind (which is something for which we do have recognized standards and judging entities). In the United States, unless you belong to one of the religious denominations whose rules about religious discourse differ from the mainstream, you'll find my behavior unseemly in the extreme, and personally embarrassing. But there is no way you can prove that my utterance either is or isn't as valid an example of religious language as, say, the Twenty-third Psalm.

Lawrence LeShan has suggested that people who don't like the mantras already available to them for meditation make their own by opening the phone book, pointing to a name, taking its first syllable, repeating the process, and combining the two syllables. "Anderson" and "Willingham" would yield the mantra anwil, for instance—and there is no way to prove that it wouldn't be as valid a mantra as any other. However unsuitable it may seem, there is no way to exclude it from the body of religious language. You may feel, as did my students, that this is a shocking state of affairs and should not be allowed, religious freedom or no religious freedom. Nevertheless, that's the way it is, and further exploration of the validity issue is the proper domain of theologians and religious scholars. The interesting question is why those who profess not to be religious should care one way or the other; they do, and they have strong opinions on the question.

For our purposes in this chapter we'll be restricting the meaning of "religious language" to sequences of language that are either customarily and typically used in religious ceremonies and practices—such as the sacred texts of a faith, like the Bible and the Koran, and their oral counterparts in many cultures; prayers and blessings and (unfortunately) curses; creeds; rituals and sacraments; and hymns—or sequences of language specifically about such items.

Attitudes and beliefs about religious language extend over an extraordinarily broad range. These attitudes and beliefs aren't just theoretical matters; they have significant real-world consequences. At one extreme is the belief that even the most seemingly trivial change in a sequence of religious language—changing a single letter or sound or gesture, for example—will ruin it. In some cases the claim is that change weakens the sequence, robbing it of some or all of its power; in others, that change would pervert the power of the sequence and make it powerful for evil rather than for good, as with the practice known as the "Black Mass." At the other extreme is the conviction that religious language is just like any other kind of language, with no special strength or power. We won't discuss the second extreme position here; I'm confident that this chapter as a whole will demonstrate that it can't be valid.

The first idea—that religious language is sacred and change is desecration—is common in Native American cultures. You will remember the example (on page 30) of the Kevesan people, who are willing to remain forever illiterate in their language because they consider the entire language sacred and feel that they couldn't control what happens to it if they allowed it to be written down. This decision threatens not only the language but the culture; the Kevesan are willing to accept that danger for the sake of their religious faith. Field-workers in many Native American languages know that when their Indian consultants agree to let them write down or record a sequence of religious language it will always contain some small change, some deliberate "mistake" or omission that makes it different from the actual sequence and thus no longer sacred. This is done not only to protect the sequence of language but also to protect the fieldworkers themselves, since the consultants believe that they can't be expected to know how to deal safely with language of such power.

I assure you that this isn't armchair anthropology, interesting but relevant only to a few very special circumstances. It has been anything but trivial for the Arabic-speaking world, where the conviction that any change in religious language is desecration has become a serious threat.

Most English-speaking people of the Western world are undisturbed by the frequently appearing new translations of the Bible. They accept and endorse the idea that this holy book should change with the times so that its message can be more easily understood by contemporary readers and listeners. We've seen the publication of a translation of the Bible into the street talk of our American inner cities. The organization Priests for Equality has published The Inclusive New Testament, in which Colossians 3:18 ("Wives, submit to your husbands … ") and 3:19 ("Husbands, love your wives … ") appear as "You who are in committed relationships … " and the begats include the names of mothers as well as fathers when the mothers' names are known.

Arabic-speaking peoples have a drastically different attitude toward the Koran, which they believe to be the actual words of their prophet, sacred in their form as well as their content. As a result, they have come to look upon written Arabic in something very much like the way that the Kevesan perceive their spoken language—as inherently sacred.

In a story in the Wall Street Journal ("War of Words: Arabic Emerges as a Weapon in Mideast Struggles"), Amy Dockser Marcus describes this problem, with comments from sources in the Arab world. "Over time," she writes, "the written language of the Koran and the local dialects spoken in Arab countries have drifted far apart. This gap has led to growing numbers of people who never master the Arabic used in books, newspapers, or official documents." She writes that Dr. Mohamed Maamouri of the International Literacy Institute wants to see written Arabic modernized so that the language taught in the schools will be closer to the language that students actually hear and speak in their homes, but he finds that "most of the region's governments remain paralyzed, fearful of being portrayed as tampering with the holy language of the Koran." Marwan Juma, general manager of Jordan's Internet service, is distressed and worried because this resistance to any kind of change in the written language means that no Arabic interface is yet available for use on-line. "Without Arabic," he says, "we face a dying heritage and loss of identity" (Marcus 1997).

What would it mean if a similar situation existed in the United States? In "A Language Divided Against Itself," Chris Hedges explores this idea, asking his readers:

What if 80 or 90 percent of Americans spoke every day in the brutal and angry cadences of gangsta rap, while the members of a feudal upper class mused over their own demise in Elizabethan English?

(Hedges 1996)

That is, what if written English had been frozen in the forms that were used in the King James Bible, while the spoken language had moved on, and the gulf between the two varieties of English had restricted literacy to an elite upper class?

Suppose politicians in the United States could take something like "Thou shalt not suffer a witch to live" from the King James Bible and turn it into a rap slogan aimed at illiterate thousands in American inner cities and rural small towns, with complete confidence that its targets wouldn't be able to find that section of the Bible and read the sequence in context, or read commentaries and discussions of the section. Suppose that at the same time the elite classes in the United States would know so little about rap that they'd pay no attention to the slogan and its effects. Suppose politicians could take the loaded word "abomination," which figures so prominently in the Old Testament, reshape it in a dialect unfamiliar to upper-class Americans, and begin using it as a buzzword to whip up hatred against whatever they disapproved of. This is precisely the sort of problem that Arabic-speaking peoples struggle with in the real world today.

Hedges acknowledges that the image is an exaggeration for many privileged speakers of Arabic, but reports that it accurately reflects the linguistic circumstances of the poor throughout the Middle East, whose inability to read written Arabic and interpret it for themselves leaves them dangerously vulnerable. Amy Docker Marcus quotes the Arab poet and songwriter Nabeel el-Khadeer's lament that "Language is the weapon of our times." And Hedges makes it clear that although Arabic may seem far outside American concerns, it in fact affects Americans directly: He explains that the carefully crafted inflammatory slogans and buzzwords aimed at the poor by politicians are "almost incomprehensible to educated Arabs, only widening a dangerous gulf between an elite that looks to the West and an enraged underclass from which suicide bombers and murderers of tourists can be recruited" (Hedges 1996; emphasis added).

The situation in the mainstream cultures of the United States is different, right? There may be a little resistance to new Bible translations, but most Americans are literate, and are broad-minded about religious language. We may feel a sense of stylistic loss when the Christmas angels' "Fear not" is modernized into "Don't be afraid"; we may feel that rendering "Let not your heart be troubled" as "Don't worry" throws away much of the beauty of the earlier version. I certainly feel that way. But nobody is ready to riot against such changes. We may be offended when people pepper their speech with "Jesus Christ" or "God damn" as their preferred casual curses, but we're not likely to attack them physically for that behavior. We may be disgusted when writers and artists take liberties with subjects that we personally consider sacred, but we're not interested in putting such people in prison for their actions. Right? Yes and no. As is true of so many language-related issues, it depends. It's not simple or straightforward.

Some examples of American attitudes toward religious language are as curious as any that an anthropologist might bring back from fieldwork at the outermost fringes. As when the local government of Kleberg County, Texas, voted unanimously in favor of saying "heaven-O" as the official government greeting, never mind the fact that "hello" has nothing whatsoever to do with hell or with the word "hell." (It's instructive, and at the same time bewildering, that this information appeared on page 25 of the December 13,1997, issue of TV Guide, without a byline and with the disrespectful title "Then They All Got Back into Their Heavenicopter and Flew Back to the Mental Hospital.")

Annie Dillard, writing in Holy the Firm, describes the "set pieces of liturgy" as language that people have managed to address to their God "without getting killed."

And Stanley Fish makes a claim that is neither obviously absurd nor obviously poetic—but obviously important, linguistically and spiritually—when he writes that "Religious discourse … cannot be unconcerned with the substantive worth and veracity of its assertions, which are in fact presupposed" (Fish 1996, page 22). That is, we can't just dismiss someone's religious language casually, because unless there is an open statement to the contrary we have to assume that the speaker believes that its various propositions are true.

It's not an accident that Fish makes this claim in an article titled "Why We Can't All Just Get Along." Many Christians who agree with Stanley Fish may find themselves in the awkward position of having to use religious language which presupposes that all the world's billions of non-Christians are doomed to spend eternity in hell. That terrible claim is presupposed by numerous examples of religious language used in Christian sacred texts, sermons, hymns, and the like. It's made explicit in the recent manifesto of the Christian Evangelical denominations in the United States, published in the June 14, 1999, issue of Christianity Today: "We deny that anyone is saved in any other way than by Jesus Christ and his Gospel. The Bible offers no hope that sincere worshipers of other religions will be saved without personal faith in Jesus Christ." That's carefully worded, but its intended meaning is quite clear. (A typical response when I ask believers how they deal with this is "I try not to think about it" plus a swift change of subject.)

In a multilingual and multicultural America this particular presupposition can have complicated consequences. I have a Buddhist friend whose first visit to an American church was his last—because it included a statement from the minister, directly to my friend, that his deceased Buddhist parents were at that moment in hell and would remain there for all eternity. I've seen a number of Southern Baptist ministers, facing a point-blank outraged question like "Are you saying that all Buddhists and Jews will go to hell?" escape by answering with "I wouldn't presume to make a judgment of that kind." The preacher who talked to my friend either wasn't that adroit, or genuinely felt a religious obligation to say what he did; if the latter reason is the explanation, I'm grateful that I carry no such burden.

To understand how religious language in English affects people's lives in the United States today, we need to remember two things as we go along.

First: The effects that religious language has on an English-speaking population are effects of contemporary English, which brings with it all the presuppositions of contemporary English.

This point is far too often overlooked or forgotten. Scholars and theologians of the English-speaking world may be convinced that they know what the words of some religious sequence in English "really" mean, based on their understanding of what the corresponding sequences in ancient Aramaic or Hebrew or Greek "really" meant. I suspect that they're sometimes wrong, and not only because millennia have gone by since the words were first spoken. For all their expertise, they have no way of knowing what English body language—especially what intonation and tone of voice—should be assigned to the translations of religious utterances from their original languages.

Only English spoken by a machine (my answering machine, for example) carries no emotional meaning. Quoted English speech is most often read silently, but we hear it in the mind's ear with some pattern of intonation attached. "If you love me you will keep my commandments" means one thing, while "If you LOVE me, you will KEEP my COMMANDments!" means something else. Not even the most learned scholars—like the members of the famous Jesus Seminar, who presume to rule on which sayings attributed to Jesus he actually did or didn't say—know which of those two English utterances, let alone the multitude of other possibilities, was intended by the original non-English utterance. If the scholars are correct in the judgments they make about such things, that's admirable. But ordinary people, which includes most of us, are even more ignorant of theology and religious theory than of linguistics. We don't experience English religious language as an English way of saying some sequence of Greek or Hebrew or Aramaic that we're able to understand in the original; we understand it as the English of today.

Much time is spent in religious discourse in America, both spoken and written, struggling to explain that when Jesus said "Be ye therefore perfect as my Father in heaven is perfect" he didn't mean perfect. "Perfect" in the original language, we're told, didn't mean "perfect"; it meant something else, something more along the lines of "mature" or "whole" or "complete" but not exactly any of those either. This is little or no help to average devout Christians, who will continue to feel that they've been commanded to be perfect.

Second: The effects of religious language are in no way restricted to effects on believers, and that's not just because believers and nonbelievers share the same world.

I've asked many people who insist that they are totally nonreligious whether they find "The Lady is my shepherdess; I shall not want" offensive. Whether they are shocked by the crucified Santa Clauses that can be found on sale in Japanese gift shops at Christmas. (Bizarre as this seems, the sort of cross-cultural blundering it demonstrates can be seen in almost every gift shop—and truck stop—in America, where the symbols of different Native American religions appear scrambled and distorted in exactly the same fashion on objects of every kind.) If it makes any difference to them whether the gay lifestyle is called "something many people strongly disapprove of" or "an abomination." If they have any opinion about whether their dog has a soul, or about the point at which a fetus has a soul. I've asked them how they'd feel if they heard an emergency room nurse say to a frantic wife in the waiting room, "Mrs. Jones, the doctors have resurrected your husband and you'll be able to see him shortly." They are ordinarily amazed at how strong their reactions to my questions are and at how much these things do in fact matter to them. Finally, I've been intrigued by the number of professedly non-religious young people I know who object to any sort of disrespectful use of the phrase "The Force."

In earlier chapters we've discussed the effects of naming some human state as a disease, as in the case of renaming menopause "hypoestrogenemia." The effects of such naming extend far beyond the individual menopausal woman and into the society around her and the society at large. Similarly, when the state of weighing more than the cultural ideal is called neither a normal variation in human shape nor a disease, but is named as a sin, the consequences permeate the culture. Not to mention the fact that the contemporary understanding of the sin in question—gluttony—is a distortion ot what the term "really" meant. Gluttony is obsession with food. When that obsession takes the form of a constant preoccupation with staying rail thin, it is as much a form of gluttony as is overeating—but contemporary American English has lost track of that meaning; making it possible for one group of gluttons to feel virtuous while considering the sorry state of the other group to be evidence of immorality or moral weakness.

The same thing holds when alcoholism or other addictions, or homosexuality, or attempted suicide, are classified as sins. People behave very differently toward someone whose state is considered proof of moral weakness and wickedness (a sinner) than they do toward someone who is considered to be sick and in need of medical treatment (a patient); the insanity defense for murderers is the prototypical example of that fact. And people who believe their own state and behavior to be sinful behave differently in this world than people who are convinced that they're not personally responsible for their condition because they are its "victims."

Everyone who interacts with a sinner—family, friends, workplace associates, teachers, helping professionals, and on down the list—is affected by the acceptance of that word from the vocabulary of religious language and behaves in ways that are shaped by that word. It makes very little difference whether they are themselves believers; the effects of the words "sin," "sinful," and "sinner" on the person they're applied to bring about changes, including legal changes, and others respond to the circumstances that are created by those changes.

Western elite culture, the culture of the highly educated professional and the academic, is just as subject to the effects we've been discussing as the cultures of our less privileged and less educated citizens; the same thing is true of cultures that are still functioning in ways that Western society set aside centuries ago.

It's not always easy to disentangle religious language effects from the effects of other factors that tend to hide the religious aspect, but the religious ones are always there. We're going to look now at one of the most obvious and easily identified examples, that of the ongoing furor about the overwhelming "maleness," in linguistic terms, of Christianity and Judaism.

Case Study: The Controversy over "Inclusive" Language

If asked in public, most people in the United States will agree unhesitatingly that the Judeo-Christian God has no sexual gender, that God is neither male nor female (nor neuter, for that matter), and that the entire concept of sexual gender has no more relevance with regard to God than being accordion-pleated or paginated would have. They're absolutely right about that. In the Judeo-Christian culture, to think otherwise—to worship a God who is understood to have the shape and form of a human being of one gender or the other—is to commit idolatry. Nevertheless, few controversies have so torn the Judeo-Christian churches apart and so roused Judeo-Christian passions in America as the ongoing struggle over the masculine/feminine domain in what is usually called "inclusive" language. (Strictly speaking, "inclusive" language means that the religious language must include ethnic groups other than whites, must include the aged and the handicapped as well as the young and able, and so on. But the dispute over male versus female linguistic items has almost taken over the term.)

English is one of the languages that makes a fuss about the specification of human biological gender; it's not possible in English to use a singular third-person pronoun without making a specific choice among masculine "he," feminine "she," or neuter "it." It's this characteristic of the language that is responsible for Rosemary Ruether's famous, flawless, and splendidly contradictory example, "God is not male; He is Spirit."

English speakers have been trying for centuries to escape from this bind by using "they, their, theirs, them" as singular third-person pronouns, as in "Every member of the congregation had their own hymnal." (Contrary to popular belief, this practice is not a recent development.) When they do that, however, they're told three things: first, that they're making a grammar error; second, that substituting "his or her" for "their" in such sequences is awkward and cumbersome and should be avoided; and third, that the correct way to express their meaning is "Every member of the congregation had his own hymnal"—because "everybody knows the word 'his' in that sort of sentence means 'his or her.'"

The first and second of these statements are subject to argument; they depend on what dialect is being used and on opinions about style. The third, however, is simply false. Research has more than amply demonstrated that people today interpret as exclusively male the so-called generic masculine items of English like the masculine pronouns and words such as "mankind," all of which are alleged in grammar books and style manuals to include their corresponding feminine meanings. When research subjects are given sets of sentences such as "Every doctor dreads his Board exams" and "The composer who doesn't play the piano is hampered in his work" and are asked to draw pictures to illustrate them, by overwhelming majorities they draw pictures of a male doctor, a male composer, and so on.

No evidence whatsoever exists that masculine vocabulary and imagery are any more genuinely "inclusive" in religious language than in nonreligious language. We have no reason to assume that if research subjects were asked to draw illustrations of sentences such as "Every sinner hopes that he will be forgiven" they would change their behavior and draw roughly as many female sinners as male ones. This is no problem in religious life for men, who find themselves everywhere included and represented. But the situation for women is different.

Theologian Elizabeth A. Johnson writes:

What results when the human reality used to point to God is always and everywhere male? The sacred character of maleness is revealed, while femaleness is relegated to the unholy darkness without … This state of affairs has a profound impact on women's religious identity.

(Johnson 1993, page 37)

It's ironic that this quotation, which so perfectly argues the religious language case for women, should bring us right up against another major problem of religious language today. For people of color the constant association of holiness with whiteness, and of wickedness with blackness—as in "the unholy darkness without"—causes great pain. The good and the holy in religious language is always light and white; we never hear that our sins might be washed as black as ebony.

Claudia Camp is even more specific:

Because the Christian traditions have for so long put "God" and "father" in the same category, this conjunction has had a real historical impact. "God is a father" is a literal statement within the discourse.

(Camp 1993, page 33)

We've run up against the linguistic relativity hypothesis controversy again. If linguist Joseph Beaver is right that "grammar and attitudes have nothing to do with each other," then we're talking about nothing more than style and taste. If Benjamin Whorf was right, we're talking about something far more important. If, as Pinker would argue, all that's necessary to clarify matters for a woman who feels excluded from Judeo-Christian religious language is to point out to her that all those male terms are "generic," we have no problem; if Pinker is wrong, the problem is very serious.

Hymn-writer Brian Wren states it very well:

If our language about God is harmless, and has no effect on our thinking and behavior, then the fact that all our names and images of God are male is unimportant … On the other hand, if language powerfully shapes our thinking and behavior, then the maleness of God-language … becomes a crucial issue.

(Wren 1991, page 55)

Over the course of more than a quarter century of investigating religious language in American English, I've asked hundreds of churchgoing Christians—in privacy, and with my promise of complete confidentiality—what they really believe about God's gender. The response, a great deal of the time, has gone like this: "I know I'm supposed to believe that God doesn't have any gender, and that's what I always say—but really, I think God is male." In some cases the answer has ended with the astonishing "I think God is a man. Not a human man, of course—but a man." What does "a man, but not a human man" mean? I have no idea. I haven't had the opportunity to do this informal experiment with the Jewish or Islamic counterparts of these Christians, unfortunately; but if I had to guess, I'd guess that the results would be similar. I have had the opportunity to ask large numbers of people who claim to be nonreligious; when they don't fend me off with "I have no opinions about that, since it's all nonsense anyway," for the most part they agree that God is "really" male.

As this would predict, those who complain about the "generic masculine" in the English Bible, buttressed as it is by an endless proliferation of explicitly male terms like "Father" and "Lord" and "King" and "Son of God," meet with intense criticism from the public as well as from professionals in religion. Zondervan (a major American religious publisher) recently brought out a very mildly inclusive Bible. It used such rephrasings as "Child of God" rather than "Son of God" and "humankind" rather than "mankind"; it changed many singular verbs to plural so that the genderless "they" could be used instead of the gendered "he or "she." The result was such a firestorm of criticism that printing was hastily stopped, despite the financial loss for the publishing house.

It's one thing when such criticism comes from denominations which openly claim that God intends for women to be subordinate to men. That is at least in part a logically consistent position. I say "in part" because so long as Judeo-Christian theology insists that God has no sexual gender, it's illogical to choose any gender at all for language that refers to that God. You don't have to. It's possible to just say "God" everywhere, without any pronouns; the result is not always elegant, but it's entirely comprehensible—and entirely logical when everyone involved is insisting that the entity in question has no sexual gender. Here's an example:

Thus, when God's Spirit "hovers over" the "face" of the earth, we learn that God's relation to creation was an intimate, loving one as a mother hen protecting her chicks, a metaphor God when on earth will repeat.

(Spencer and Spencer 1998, page 23)

Ruether's "God is not male; He is spirit" could simply be "God is not male: God is spirit," and the contradiction would disappear. Even "Godself" for "Himself," if it were routinely used, would soon sound normal and ordinary. But doing that would mean deliberately changing existing religious texts, and the resistance to that—in the United States just as in the Arab nations—is very strong and deeply entrenched.

Ordinarily, individuals who insist that the religious language of Judeo-Christianity must not be changed—either by doing without pronouns for the divine or by making it "inclusive"—also insist that (a) they know very well that God has no sexual gender, and (b) they know very well that men and women are equal in God's sight and should also be equal in the sight of man (by which they of course claim to mean everyone, both male and female, and transgendered).

Suppose we assume that they do in fact believe exactly what they claim to believe. If so, it requires cognitive and linguistic gymnastics of the most intricate kind, and a consistent strategy of reframing, as in the recent very clever editorial by David Neff in Christianity Today which claimed that the argument has nothing to do with sexual gender anyway. Gender, he argues, is a red herring. His title is "The Great Translation Debate," and his claim is that the entire inclusive language argument is not about gender at all, but about how to be "most faithful to the ancient text" (Neff 1997, page 16). Or as stated bluntly by Grant R. Osborne in "Do Inclusive Language Bibles Distort Scripture? No," which opens with this sentence: "Whether or not to use inclusive language in Bible translation is not a gender issue but a matter of translation theory" (Osborne 1997, page 33). That may be true for scholars like Osborne, but for the ordinary person in America it is a gender issue.

We need to be aware that even those who are most severe in their insistence that the Bible's language be translated with utmost fidelity make exceptions in some cases. For example, there is Jeremiah 20:7. In the King James Bible it reads, "O Lord, thou hast deceived me, and I was deceived; thou art stronger than I, and hast prevailed; I am in derision daily, everyone mocketh me." The New King James has it as "O Lord, You induced me, and I was persuaded; You are stronger than I, and have prevailed. I am in derision daily; everyone mocks me." Scholar Abraham J. Herschel gives the literal meanings of the two verbs in question as "wrongfully inducing a woman to consent to prenuptial intercourse" and "violent forcing of a woman to submit to extranuptial intercourse." He offers the translation "O Lord, Thou has seduced me, and I am seduced; Thou hast raped me, and I am overcome." This would certainly make the clause about being mocked a good deal easier to understand. But even the liberal Jerusalem Bible, which starts out with "You have seduced me, Yah-weh, and I have let myself be seduced," isn't willing to go that far; it continues with "you have overpowered me; you were the stronger. I am a daily laughingstock." Faithful translation may be the professed goal, but when the result would have Jeremiah accusing the Deity of having raped him, the literalists back down. Clearly, there are limits to this professed goal of "being most faithful to the ancient text."

Here are three more examples of commentary about inclusive language, to show you how varied and ingenious they can be, and how impassioned:

Even if speakers/writers should self-consciously choose not to use the masculine pronouns for God, the fact remains that they are self-consciously choosing not to use them; they are intentionally deciding to go against internalized grammatical convention and the deep, unconscious habits of the society … Thus the practice of avoiding the masculine pronoun when talking about God will never become habitual and routine; such practice must and will remain at the level of self-aware rebellion.

(Hook and Kimel 1993, pp. 16-17)

While women, in their own fashion, may sin as much as men, it is the pride and willfulness and violence of men that most radically disrupts the world's peace and order. Feminist reformers may have no trouble in recognizing this fact, but they seem not to notice that breaking the arrogance and power of men requires a God who is, among other things, the Judge of Nations, the Lord of Hosts, the "King of glory … mighty in battle." … God's lordship is the only plausible check on the predatory tendencies of wicked men and nations, from Pharaoh to Haman to Hider.

(Berke 1996, page 35)

The most simple and compelling reason to reject gender-reimaging is that God, though not a man or a male, has used masculine terms in making His self-disclosure to humanity.

(Berke 1996, page 37)

The second Berke quotation, if taken seriously, would lead us straight to the isolationist position held by Arab speakers toward the Koran, in which the words in the religious texts are perceived as literally the words of God Godself. And how would we then carry on the evangelical practice of translating the Bible into the many human languages in which there are no masculine pronouns?

It might seem that all we'd have to do is make sure that a new pronoun, an exclusively masculine pronoun, is added to those languages that don't already have one. However, we have no reason to believe that that would be any easier to accomplish in other languages than it has been in English. There have been nearly 100 attempts in recent history to give English a gender-neutral third-person pronoun to avoid the constant "he or she" problems, and every single one of those attempts has failed. This is typical linguistic pathology; everybody, in their daily speech, uses "they" as a way out of the quandary; everybody, when talking about what is "correct," says it's wrong to use "they" to refer to just one person.

It's easy to add a new name-for-a-widget to a language; we do that for English almost every time something new is invented or discovered. It's also easy to make changes that do nothing but substitute one widget (as long as it's not a sacred Widget) for another. The English Bible says that marriage means a man shall leave his father and mother and cleave to his wife. Bible translators Gene and Marie Scott translated that into Sharanhua as "For this reason a man shall have the same mosquito net as his wife"; the Sharanhua of Peru don't leave their parents' homes when they marry. (See Zoba 1997, "Your Sins Shall Be White as Yucca," for details.) This sort of thing has to be done all the time in translating the Bible. But adding a new pronoun to a language is different.

Pronouns carry powerful presuppositions about reality as the speakers of that language perceive it. It matters to English speakers whether something being spoken or written about is animate or inanimate, human or nonhuman, one or more than one, male or female or neither. Sentences that lack such information are totally unacceptable. Speakers of other languages are equally unable to imagine using pronouns that make distinctions differently—that divide up reality differently—than the way they're accustomed to doing it. Pronouns are one of our most powerful tools for specifying what things members of a culture must pay attention to, and there will always be strong resistance to their modification.

We're often told that all of this controversy is a tempest in a teapot because, since not one of us has ever perceived what we are trying to describe and discuss, all talk about the Holy One is by definition metaphorical. That's all very well in theory, but people don't understand beloved religious metaphors as "only figures of speech." Trained religious professionals such as clergy and monks and nuns may be able to ignore the symbol and worship only the Divine Reality behind it; very few ordinary people are able to do that. We may know intellectually that our Heavenly Father is an unimaginable disembodied genderless divine Something-or-other; but when we pray to that Heavenly Father the image in our minds and hearts is probably a very human-looking man, something like Abraham Lincoln or Paul Newman, or their equivalents in other ethnic groups.

We could easily test the hypothesis that speakers of American English really perceive the Christian God as having no sexual gender. We would only have to produce a new biblical movie in which God speaks from the burning bush—and the voice of God is neither Charlton Heston nor James Earl Jones but the voice of a woman. Diane Sawyer, perhaps, or Jessye Norman. We all know what would happen, however much we claim to understand and believe that God is genderless. The people in the audience would react with a shocked silence, perhaps a shocked gasp. And then, depending on how devout they were and what denomination they belonged to, they would either burst out laughing or stand up and march out of the theater in outrage.

Male … Female … and?

I have restricted the discussion in this section to the two traditionally recognized biological genders: male and female. That has religious-language precedent; the Bible says "Male and female created He them." There is today, however, a vigorously thriving academic and scientific field known as "gender studies." The field of gender studies proposes that it's an error to assume automatically that a human being is either male or female, with no other possibilities to choose from. It also claims that sexual gender is not necessarily a biological matter, but is at least in part the result of socialization. If the gender studies scholars and scientists are correct, the traditional categories—and the pronouns associated with them—are not enough. If English is not willing to move to a genderless third-person pronoun, they would tell us, then it will have to find additional pronouns to refer to those human beings who insist that neither male nor female pronouns accurately name their gender.

This book isn't an appropriate forum for taking up this argument. It will be obvious, however, that adding more categories for gender would only make the inclusive language arguments even more difficult and more passionate than they are when we have only "he" and "she" to deal with. It would be far simpler, far more logical—and unquestionably more consistent with the theological precept that God is genderless!—if we eliminated the biological gender distinction from our religious language altogether. We can be certain, however, given the ferocious opposition to making such a change even when only the male/female distinction is involved, that trying to do it for the sake of additional genders would be impossible.

The inclusive religious language controversy revolves around a metaphor, obviously. That metaphor—GOD IS A FATHER, which presupposes maleness and dominance—has been a source of comfort to humankind through the ages; it has also been, and is still, a source of conflict and division and turmoil. It can't be a definition, like "A bat is a mammal," and subject to scientific objective testing; it can only be a metaphor.

"Fine," says the practical person. "Since we know that, let's just get rid of it. Better yet, let's get all the metaphors out of religious language!" It can't be done. This is one area of human language in which metaphor is our only resource. We can perhaps talk about ethics, and about morality, without metaphor, but religious language without metaphor is impossible.

Other Metaphors in Religious Language

Another part of life in these United States where religious language has profound effects, and where those effects extend beyond believers, is in our choice of a unifying metaphor around which we can organize our lives and our behavior. Because roughly half of our population is female, the metaphor of the Father won't serve that purpose, and trying to switch to the Parent would land us right back in the inclusive language mess. Something else has to be selected.

Christians (who are the majority religion in the United States) read in the New Testament that Jesus has commanded them to return good for evil; they read that it is the meek and the peacemakers who will inherit the earth; they read that the rich and the powerful will have great difficulty getting into heaven. (And much more of the same.) They sing hymns about "Gentle Jesus, meek and mild"; they memorize and love the Twenty-third Psalm with its description of the Lord as a shepherd. All admirable in theory; all downright pretty, like the Christmas cards where the lion and the lamb lie peacefully cuddled up together under a tree.

But when the time comes to choose a metaphor around which to organize their religious/spiritual lives, in the way that we've chosen the metaphor of the Football Game as an organizing mechanism for our business lives, no gentle shepherds need apply. Even the quickest glance around the shelves of a Christian bookstore will make it unmistakably clear that the dominant unifying religious metaphor for the ordinary religious person is the Warrior.

Warriors don't turn the other cheek, they don't return good for evil, they don't give the thief who steals their shirt their coat to go with it. Warriors get out there and do just one thing: They fight to win.

For the adults fulfilling this role, the script for this metaphor is clear. If they're in combat, if they're mentally ill, if they're fanatics, or if they're in contact sports, they may do some of their fighting physically. Otherwise, in a society such as ours that has legal penalties (not always enforced, but at least on the books ) against physical violence, they do their fighting verbally. Our culture honors and rewards them for this behavior. Unless they become famous or accomplish something truly remarkable, as did Martin Luther King, Jr., we call people who return good for evil "wimps" and "weaklings."

I have been flabbergasted by this my entire adult life, and awed by it. It's particularly baffling to me that we see this in its most extreme form in those devout persons who claim to believe that the Bible must be taken absolutely literally; no one buys more books and tapes and videos about "spiritual warfare" than the biblical literalist. Only among such groups do we find the extraordinary current practice of holding prayer meetings at which you pray for misfortunes or tragedies to happen to those you oppose, like the groups whose admitted purpose was to pray for the death of President Clinton.

What are the effects of this metaphor on the culture at large? It has turned us into a people who, while paying lip service to peace—and reacting with horror to extreme violent events such as the Oklahoma bombing and the sequence of school massacres—allow our lives to revolve around violence. We are a people who spend large portions of our disposable incomes supporting violent movies and books, who allow our children to spend hours of their day playing violent computer games, who do nothing more than shake our heads when we learn that one product available to our children is a set of Serial Killer trading cards, and who—inevitably—conduct our business by the rules of football rather than the rules of the Bible, using ruthless competition to win as our governing strategy. Those who profess to take the Bible literally and want it applied literally in many areas of human life—especially the area of sexual behavior—very rarely want to extend that idea to business and economics, where it would mean that they were forbidden to buy and sell real estate, forbidden in most cases to lend money at interest, required to return good for evil, and much more of the same kind. As Walter Wink puts it:

The myth of redemptive violence is the simplest, laziest, most exciting, uncomplicated, irrational, and primitive depiction of evil the world has ever known. Furthermore, its orientation toward evil is one into which virtually all modern children (boys especially) are socialized in the process of maturation.

(Wink 1998, page 53 )

I agree with Wink's statement; I would add only that the "boys especially" in question are for the most part brought up by women; little girls must therefore be socialized to be willing to participate in this process. By "redemptive" violence, Wink is referring to the concept that violence is acceptable, even admirable, when its goal is to accomplish good, as in just wars.

The strength of the Spiritual Warrior metaphor is growing daily, and it has spread from the domain of actual combat to the more pervasive domain of all competition. It used to specify only that the Warrior must win; now it is understood to demand that everybody else except the Warrior must lose. That is, there can be only one winner, and that winner must leave an empty battlefield behind. We saw this in the recent Olympic Games when the Nike ads at Atlanta said, "You don't win silver. You lose gold." And our young people, who've come into this semantic controversy late in the battle—or in the third quarter—and are not being taught the relevant history, have lost track of the principle of the just war. That principle stipulates that violence must always be the last resort, carried out with heartfelt regret because everything else has been tried and has failed, and no other action appears to be possible. The Warrior metaphor has brutalized our sports and is brutalizing our lives, and it is a religious language phenomenon.

The most extraordinary section in Lakoff and Johnson's Philosophy in the Flesh (1999) is pages 563-564, where they claim that—having proved to their satisfaction that there cannot be a disembodied mind—they extrapolate from that to the claim that there cannot be a disembodied soul. That is, there cannot be a mind with an existence independent of the body; therefore, there cannot be a soul with an existence independent of the body. They do not even consider the possibility that a divine Creator would be abie to construct something—something which human beings could refer to as "the soul"—and give it characteristics so different from the mind that it would be entirely capable of independence from the body.

The greatest failing of the melting pot metaphor is that it overreaches. It exaggerates the degree to which immigrants' ethnicity is likely to be extinguished by exposure to American society and it exaggerates the need to extinguish ethnicity … Perhaps a new assimilation metaphor should be introduced—one that depends not on a mechanical process like the melting pot but on human dynamics. Assimilation might be viewed as more akin to religious conversion than anything else. In the terms of this metaphor, the immigrant is the convert, American society is the religious order being joined, and assimilation is the process by which the conversion takes place … Conversion is a mutual decision requiring affirmation by both the convert and the religious order he or she wishes to join. Converts are expected in most (but not all) cases to renounce their old religions. But converts do not have to change their behavior in any respects other than those that relate to the new religion. (Salins 1997, page 23; emphasis added)

That is, someone who has openly joined a Christian church in the United States must sit quietly and respectfully, without objecting, while a minister reads and preaches about turning the other cheek and following the Golden Rule, and even reads verses of the Bible commanding us not to charge interest and not to bring lawsuits—but once that person leaves the church, no one expects that the Shepherd will replace the Warrior in his or her life. Perhaps, if we treat assimilation like conversion, as long as those who assimilate observe all the rules of public life and public citizenship there's no reason in theory why they couldn't continue to maintain their ethnic heritage and their native language in the privacy of their homes. The assimilated multilingual individual would end up with a string of beads perhaps even less matched, even less a harmonious whole, than usual—but they would remain just one string of beads.

In theory, this sort of lip service model for assimilation is appealing; in religious life, it might even work. As always, the string of beads would have to be left at home when the multilingual went off to work.

A Final Note—Religion plus Music

Throughout this book I've repeatedly pointed out the importance for English of the melody attached to words by tone of voice and intonation of the voice—the tune the words are set to. In most forms of discourse that is as melodic as it's possible to be; only in musical comedies do adults burst into song in the middle of their conversations and other language interactions. We may sing lullabies to tiny kids, we may sing in the shower or in the car, but that's as far as it goes.

If you and I are sitting around talking about religious matters, we're still not likely to burst into song. But for churchgoing Americans, the singing of religious language in the form of hymns—not just listening, but actually singing along—will be a regular feature of life at least once a week. It would be remiss of me not to mention this, and not to point out that language set to music with a pleasing melody has a power that spoken language rarely can achieve. People will listen, and listen willingly, to sung words that they wouldn't tolerate for a minute if spoken—something that the writers of "protest songs" are well aware of. People (even totally nonreligious people) who would leave immediately if others started praying aloud at them will happily sit and listen to the singing of religious music. People who have little interest in a cause will unite around it because of the extraordinary attraction of a song—we only have to think of what was accomplished in our past by Julia Ward Howe's "Battle Hymn of the Republic."

Brian Wren has used this aspect of language, combined with his great skill in the constructing of metaphors, to remarkable effect. There is his hymn "Holy Weaver, Deftly Intertwining," in which he carries the metaphor GOD IS A WEAVER through the verses. There's "Joyful is the Dark," in which he turns the usual EVIL IS BLACK/ GOOD IS WHITE metaphor pair around, using phrases such as "the roaring, looming thundercloud of glory." In "Bring Many Names" he explores a set of metaphors, including not only "Warm father God" but also "Strong mother God" and "Old, aching God, grey with endless care." Songs can build bridges that spoken words cannot, and that speakers can only long for and envy. There are many reasons for this in theory; in practical terms, the reason is simply that people are more willing to listen to another person sing than they are to listen to another person talk.

So what's the solution? "Thoughtfully planned bilingualism" says Einar Haugen, a linguist … Haugen wants each of us to have a "native, homely, familiar everyday language in which we can live and love" as well as a second "language of wider communication" that will enable us to jet around the world.

(Vines 1996, page 27)

Remember Einar Haugen's proposal, mentioned previously on page 35? ***

Is Haugen right? Maybe. When the communication problems of a multilingual world are discussed, people often express agreement with him (or with some variation on his idea) and suggest that an obvious solution is staring us right in the face: Let's have a single language (traditionally called an "international auxiliary language," or IAL) that everyone would learn in early childhood, so that all human beings on this planet would have at least one shared medium for communication. What could be more logical, or more practical?

Suppose that, for purposes of discussing this idea, we call our hypothetical IAL "WorldSpeak." We'd need to call it something far less English and far more internationally neutral in form, and choosing the final name would have to be done with great care and attention, but WorldSpeak will do for our present purposes. The scenario that carries out this idea and makes WorldSpeak a genuine IAL—in two versions—is easily imagined.

In both versions, all children would acquire their native tongues in the usual way, along with whatever other language or languages were present as input in their language environment; this would involve no change from our present practice.

In both versions, all adults would know WorldSpeak, which would put an end to the need for translators and interpreters, either human or electronic, in every aspect of international interaction. Translation and interpreting would become activities done as art rather than as necessities. Monolingual persons would exist only in the increasingly rare circumstances of peoples truly cut off from modern education and technology. This would save huge amounts of time and money and effort, freeing up resources that could be used for other purposes. Not only because of the logistics-related savings, but also because we would have fewer of the expensive international misunderstandings that result from language differences. Professional translators and interpreters (other than literary translators), along with professionals involved in machine translation, would have to go through the sort of dislocation that occurs when any work role becomes obsolete in a society; they would find other ways to make their livings, in due time.

WorldSpeak Scenario A

When children started school, WorldSpeak would be a new "basic" subject with the status of reading and writing and math—from the very first day, everywhere in the world. This would establish a global pidgin. Then, once the first generation of WorldSpeakers had children of their own, every home on the planet would be a multilingual household. And human children would start acquiring WorldSpeak by being exposed to it in the home from birth, making WorldSpeak a worldwide native language. People would then take WorldSpeak courses in school just as they now take courses in their native languages, not to learn the language but to polish their comunication skills and learn "proper" grammar and usage.

WorldSpeak Scenario B

In this version every child would learn WorldSpeak as in Scenario A, but instruction would begin much later, probably as a required subject for two semesters in the final year of high school (or its equivalent in other societies). The core inventory of WorldSpeak skills—the ability to carry on a simple conversation, read and write nontechnical materials, and do survival math—would be a requirement for graduation and for getting a job. It would be taken for granted in the way that literacy is taken for granted in the industrialized world today. Whether an individual went on beyond that level of fluency would be a matter of individual taste and need.

People involved in global business or law or diplomacy, or in any of the sciences, would routinely take additional WorldSpeak courses as part of their university and professional education. People planning a vacation in a foreign country would brush up their World-Speak skills with a television or computer course, or by studying a "WorldSpeak for Tourists" book, the way people now brush up on their high school French before spending a week in Paris. There would undoubtedly be a horde of "WorldSpeak consultants."

Just as we now assume that any adult we meet will be able to read and write and do simple math, we would assume that he or she would be able to get by after a rough fashion in WorldSpeak. There would of course be times when our assumption turned out to be incorrect, just as today we sometimes run into an illiterate adult, but that would be rare. The degree of WorldSpeak fluency would vary from person to person. For some people, especially those who had grown up in households where parents chose to use the IAL regularly in the home, "knowing" WorldSpeak would mean native or near-native fluency. For others it would mean the sort of fluency a person today typically has after two semesters of a foreign language in high school if a moderate effort is made to maintain that level of fluency in adult life.

You'll see at once that Scenario A is the simpler and more efficient choice. Things would be a bit rocky for the first generation, and there would be unanticipated glitches to work out. But from then on, with WorldSpeak having become a native language for everyone, the process would go smoothly. Scenario A takes advantage of the scientific facts about language acquisition. We know that if we want our kids to have real fluency in Chinese they need to start learning Chinese as early in life as possible, preferably as infants in the home; the same thing would be true for fluency in the international auxiliary language. There is absolutely no question about this; once the transitional period was over, Scenario A would be simpler than Scenario B in every way and would yield superior results. However, it is by no means certain, or even probable, that Scenario A would be our choice.

Think about it, please, in terms of the questions that we have been exploring in previous chapters of this book: What would it mean for the human race if it were not only the case that monolingualism had disappeared from this earth, but also that every human being shared a single native language? We'd be rid of the expense and inconvenience associated with translating and interpreting, yes; international affairs could be conducted more efficiently and swiftly, yes. But what other consequences might there be? We'll come back to this issue later in the chapter.

Once the seemingly logical and overwhelmingly obvious decision has been made that an IAL would be a good thing and that efforts should be made to put one in place, you come to the hardest and most predictable question, the one that must be settled before any other steps can be taken:

Which language is going to be the one that every human being will be expected to know?

The natural human reaction is an immediate "Well, my language, of course!" but we are well aware that it's not that easy. There are only two possible answers:

• We choose some existing natural language and see that every educated person knows it, in the way that every educated European person was once expected to know Latin.
• We choose a language that's not a natural language but has been constructed (often for the purpose of serving as an IAL, in the way Esperanto was intended by its designer). We could pick one of the hundreds of such languages that have already been proposed, or we could construct a new one (in which case one can only hope that linguists would be consulted).

In either case, we could modify the chosen language to produce an "international version" of it if we wished. The most likely modification would be to simplify the language in arbitrary ways, as was done for Charles K. Ogden's 850-word "Basic English" or the version of English called "Simplified English" that all airplane pilots and air traffic controllers are today expected to know. Finally, although they may seem unlikely choices to most of us, we must not rule out in advance the idea of a sign language, or of systems such as Morse Code, when making our decision. They might have advantages that don't come immediately to mind. (A system now exists for writing and typing sign languages, called Sign Writing; it's cumbersome, but it works.)

Let's consider those two basic alternatives: natural language versus constructed language.

Choosing an Existing Natural Language as the IAL

It's disconcerting, because none of us has ever been allowed to make our wishes on the subject known, but it has to be admitted: This is already under way. In a very informal and disorganized fashion, the world has already chosen English (or, as speakers of many other languages would insist, has had English imposed upon it) as its global language. If that's the wrong choice, either because some other natural language would be better for the purpose or because we should instead choose a constructed language, we need to recognize that in a hurry, while there may still be time to stop it.

A few years ago Pakistan set up a language study commission, which concluded that "there is no escape for any country in the world from learning English well, and thoroughly" (Reeves 1997). When South Africa ended white minority rule in 1994, its official languages were Afrikaans and English; the new government added nine African languages spoken in the country to the official language list. The result? "Faced with making preflight safety announcements in 11 languages, South African Airlines pared down to English. Similar decisions have given English near supremacy at universities, Parliament, the post office and the telephone company" (Reber 1998). In a February 21, 1999, on-line review of Manfred Gerlach's book Even More Englishes, we read that it's a collection of papers "devoted to general problems of the world language" and that two papers investigate "the form and functions of the world language" in South Africa and Nigeria.

"Speakers of other languages," writes Gordon Bilney from Australia, "won't necessarily think it's fair that they should have to learn English to get the most out of life in the 21st century; nor is it, but there's little choice." (Bilney 1998.)

There's no way to avoid it. We have to acknowledge that while American politicians are fretting about the perils of bilingual education—and making speeches warning us that without "English Only" laws we face possible secession attempts by Florida and New Mexico, as Canada faces possible secession by Quebec—English has already, literally, begun taking over the world.

Let's assume for a moment that this situation is going to continue. If English is to be the final Terran choice, attempts will undoubtedly be made to bring a degree of order into the process. In spite of the dismal track record of all "language academies" and "bureaus of language" throughout history, we will surely try again. As Barry Newman points out, with obvious uneasiness, "English no longer simply divides itself into regionally rooted dialects—or distinct 'Englishes' as some argue—but into a rabble of non-native registers … for specific purposes … What happens to this language is no longer our prerogative. English is no longer our possession" (Newman, 1993). Oops!

On May 16, 1996, an announcement was posted on the Internet by Macquarie Limited's Sue Butler, for a series of international conferences. Their unifying topic was "English as an Asian Language"(a later slip of the key said revealingly that English "is an Asian Language"), beginning with a conference on Thai English to be held in Bangkok and one on Philippine English to be held in Manila. Definitions of terms included these two very significant items:

A national variety of English is a form or dialect of English which has a national distribution. It is acknowledged to be the form of English identified with the particular nation, having developed features which distinguish it from other varieties of English as a consequence of its separate history, geographical location and cultural context.

Standard English should be defined as that accepted form of English which occurs in prestigious English-language newspapers, publishing and broadcasting in the specified country.

Separate history? Separate cultural context? A standard based on "prestigious" media output in "the specified country," never mind how English might be used in the New York Times? Oops!

It doesn't help matters that we have no accepted definition of Standard English that could be proposed as an international standard. Existing definitions are of two kinds. In one group you'll find a multitude of safely vacuous variations on "Standard English is the language used by educated people." (That is, "We know Standard English when we see it or hear it.") The other group contains dangerously specific definitions that betray the ongoing public ignorance of basic linguistic facts, a situation for which the public can't be blamed but which desperately needs repair. Consider, for example, this definition from the state of Virginia's House Bill No. 2437, passed not in 1900 but in 1997, with emphasis added by me:

Standard English includes the written and spoken language which is accepted by generally recognized authorities as grammatically correct in the United States and shall not include any dialect, jargon, patois or vernacular based on the English language.

Given this state of affairs, it's reasonable to wonder what will become of English as it spreads like some variety of linguistic kudzu into every corner of Earth; but that's something that we know, from experience, cannot be predicted. It's impossible to keep a living language from changing, although that won't keep the governments of Earth from trying.

Something of this kind perhaps happened at that point in the mists of prehistory when everyone in the known world spoke a single language—let's call it Ur. Although there were no media, prestigious or otherwise, there were surely standards. We can imagine one Early Human sneering at another's inferior way of speaking Ur. And then millennia went by, and instead of Ur there were French and Albanian and Chinese and Basque and English and Mohawk. But our world is not the world that Early Humans knew. In a world that has satellite television and the Internet, and telephones that will work in the middle of the most barren desert, human beings who are separated by geography are no longer isolated from one another linguistically. Only a planetwide catastrophe that restored prehistorical conditions could once again make it even remotely possible for words to change gradually over time, in isolation, to such an extent that only highly trained specialists in historical linguistics would be able to tell that they had ever been the same. What happened to Ur (or conceivably to several "Urs") could not happen to International English.

Well, then, can we say that the whole issue of IALs is now moot? Is it settled that WorldSpeak will be some impossible-to-predict form of English? Should those of us lucky enough to already know one or more varieties of English just relax and be grateful that we happened to be born into an English-speaking language environment?

Nobody knows. The Englishing of Earth is something like the global warming of Earth. It may be happening; some experts say so. It may already be an accomplished fact; it may be too late to stop it; some experts say so. In which case, instead of passing laws to protect English we should be doing our utmost to protect our remaining store of other languages! We know from the biological sciences how dangerous it is to reduce any pool of existing varieties to just one. Nothing we know in linguistic science gives us reason to believe that it would be any less dangerous for the world's language "gene pool" to be reduced to only English, even with a multitude of dialects.

Furthermore, before native speakers of English jump to any conclusions, they would be wise to ask native speakers of other languages how they feel about the idea that English should be the lingua franca (which meant "French language") for today's human race. And then they would be wise to back off a good long distance to spare their ears after asking. Many peoples in this world—perhaps almost all—deeply resent the idea that they should sit back and allow all international commerce and science and diplomacy, all international communication, to be conducted in English. A recent attempt by an Internet service to require all communication about soccer to be in English—a very small area of Earthly communication—provoked so passionately negative and furious a response that it was hastily canceled. Even nations whose dominant language is English are fighting to keep their own brand from going under; the Sydney Morning Herald for August 14,1997, reported that a group of Australian researchers have "produced an Australian weapon to fight … American linguistic imperialism," in the form of a dictionary of Australian words and phrases to be substituted for American linguistic imports (Jopson 1997).

Consider the program now being proposed by the foremost linguist of France, Claude Hagege, in his 1996 book L'Enfant aux Deux Langues. H. Stephen Straight (reviewing the book in the March 1998 issue of Language) says on page 140 that Hagege "openly addresses what may well be the most pressing question of educational policy today: how to prepare French schoolchildren—and by extension schoolchildren across Europe and around the world—for the multilingualism that will pervade the 21st century" (Straight 1998, page 140; emphasis added). And then, to the question of which languages should be taught in French elementary schools:

Hagege heart-stoppingly responds—if I may coin an acronym—ABE, Anything But English. The world-wide popularity of English threatens to divert French school-children from the task of developing proficiencies in the many other languages of the European community and of the world.

(Straight 1998, page 140)

Anything but English! It would fit nicely on a bumper sticker or a coffee mug, and it contrasts nicely with English Only. Hagege's proposal is that the French elementary schools offer German, Italian, Portuguese, Spanish, and (for speakers of minority languages in France) French, followed in high school by two additional languages. The goal is that each French youngster would know four languages: the native tongue, two additional European languages learned in early childhood, and one other language learned in adolescence. Only at the high school level would English be one of the possible choices, and even then it would have to compete with Russian, Arabic, Hungarian, Swahili, Chinese, Hindi, Japanese, and more. Hagege perceives this project as an urgently necessary measure for preventing the global linguistic takeover of English. It's quite clear; speakers of other languages may not be able to prevent it—it may be too late to stop the English juggernaut—but they resent it from the bottom of their hearts.

The fact that English appears to be the easiest solution to the problem of global Babel doesn't mean that it is necessarily the best solution. Perhaps a serious effort ought to be made to call it off. Native speakers of English, because they suffer from the same native-language bias as any other human beings, are the last group who should sit in judgment here. We have to ask ourselves: "What if my native language were Gaelic? Or Arapaho? Or Maori? How pleased would I be with English as an IAL?"

Innocent though we native speakers of English may be personally of any interference in the affairs of other nations and nationals, the political facts are real and have to be faced: For most of the world, English is the Superpower Language. It is the ultimate "colonial" language. It's the language of the Ugly American and the Snobbish Brit. It's the language callously imposed on the downtrodden immigrant or subjugated indigenous population. For the sake of peace in the world, perhaps we should set aside our natural preference and our gut feelings and consider other alternatives.

The first "other alternative" is that we should choose some other existing natural language instead of English. But how do we do that? Many, if not all, nations would insist that their language should be the one chosen, and would offer a strong case. The French would point to their literature, their history, the tradition of French as a lingua franca, and the fact that French uses the Roman alphabet; other European nations would present similar arguments. The Japanese (and other Asian nations) would point to their own proud literary tradition, their history, their position in the world, and the fact that their writing system does not carry the stigma of Western colonialism as the Roman alphabet does. The Hopi (and many other Native American nations and indigenous peoples around the world) would insist that it is precisely the apparent difference between their language and the more commonly used languages, not to mention the absence of a recent history of conquest, that would make Hopi the ideal choice.

And so it would go, with all this passionate wrangling being carried on through translators and interpreters. It's not a prospect that gives one hope, and it's the sort of thing that has led to wars in humanity's history. Perhaps we should abandon the idea of an existing natural language as IAL altogether. Which would bring us to—

Choosing a Constructed Language (Existing or New) as the IAL

The traditional hypothesis has been that because a constructed language is the language of no nation or ethnic group, it would be free of the political problems that all natural languages bring with them. Esperanto materials frequently claim (incorrectly) that this is true of Esperanto. A distinction is usually made between auxiliary languages (auxlangs), designed with international communication as a deliberate goal, and "conlangs," usually constructed for other purposes. (The Elvish languages showcased by Tolkien in his epic Lord of the Rings and the Klingon language constructed by linguist Mark Okrand for the Star Trek television series are conlangs rather than auxlangs.)

Constructed languages aren't a new idea, nor did they begin with Esperanto, and there have been hundreds of them. The first one for which we have any historical records was Lingua Ignota, a language devised by Hildegard of Bingen. We know almost nothing about it except that it existed and that—since it was constructed in an era when Latin was the recognized IAL—it was probably not meant to serve as an international tongue. The first for which we have any useful information came from one George Dalgarno of Scotland in 1661. The first that any number of people actually used, both in speech and in writing, so far as we know, was Volapük, the work of a German priest named Johann Martin Schleyer, in 1880. Volapük was so successful in Europe and the United States that by 1889 it was said to have 200,000 aficionados and 300 supporting groups. There were Volapük publications and international Volapük conferences. But the language was very difficult to learn, and it died almost as fast as it had become popular.

When the far less intimidating Esperanto, the creation of ophthalmologist Lazar (often called "Ludwig") Zamenhof, appeared in 1887, it quickly took over the territory that Volapük had been holding; it continues to this day to be the most successful of all proposed IALs. Estimates of the number of Esperanto speakers range from as low as 50,000 to about two million; if even the smaller of those numbers is accurate, it's an impressive total, far larger than the total number of speakers of many natural languages. Esperanto has a sizable literature, both original and in translation. It has a sturdy international support structure. It has been around long enough to undergo minor reform, and there are various proposed "new and improved" Esperantos. In the early 1900s, with the great linguist Otto Jespersen (himself the author of a constructed language called Novial) leading the charge, it appeared that a multinational committee and academy might actually give some version of Esperanto an official standing—but two world wars, and the dreadful internal power struggles within the IAL movement, put an end to that effort. Nothing on the same scale has been attempted since.

A language called Solresol, based entirely on the names for the seven units of the Western musical scale, was introduced by Francois Sudre, a French scientist. All its words were combinations of the syllables (do, re, mi, and so on) used to teach singing. Solresol had the unique advantage of being a language that could be sung and played and hummed and whistled as well as spoken, and for a while in the 1800s it had a considerable popularity. At one point, the French military even considered adopting it as an auxiliary means of communication.

There is clearly a very strong species-specific language instinct, or what Chomsky would call a language acquisition device, a LAD. But equally one needs a LASS, as Jerry Brunerputs it, which is a language support system; people who speak language and who encourage the use of language, and so forth.

(Oliver Sacks, quoted in Mcintyre 1995, page 95)

Well, yes—but then, on the other hand, no!

I agree with Dr. Sacks that both a LAD and a LASS are needed for the acquisition of language, but he's much braver than I am. Even for that very limited claim—that language learning requires both nature and nurture—I would hesitate to use the word "clearly," because almost nothing in the current literature on this subject can be considered clear. Even when the statements seem to be uncontroversial, they still cannot accurately be described as "clear."

In fact, any prudent individual writing on "language and the brain" today finds the abundant deniability resources of English extremely seductive. The temptation is to write everything in Hyper-deniability Mode, something like this:

Numerous hypotheses have been put forward, and a variety of proposals have been made. It has been claimed that X, but that has been contradicted forcibly by the claim that Y and the evidence for Z. No final decision among these competing and contradictory theories appears to be possible at this time. [Blah, blah, blah … ] What is said below, therefore, should not be construed as constituting the taking of a position of any kind.

English allows me to do this. You'll notice that it even allows me to avoid admitting that I am the one who insists on remaining free to deny that I've said what I am about to say. I'm tempted, but I'm not going to do it that way. The buck will stop here, but with two preliminary warnings.

First: I have never dissected a single brain, and I could not interpret a brain scan even at gunpoint; I cannot, therefore, speak or write with personal authority about human brains in neurophysiological terms.

Second: When we hear or read anything about the human brain any less concrete than a specification of its physical weight and size—and when we read anything whatsoever about the human mind—that information must be taken with a large grain of salt. When something of that kind strikes you as plausible, the rational course is to assume that it's valid as a metaphor and use the metaphor for whatever it turns out to be worth—but leap to no other conclusions without solid supporting evidence. Lakoff and Johnson tell us in Philosophy in the Flesh that there's no way to even talk about the mind, or the mind as "part of the brain," except in metaphors.

Independent of these metaphors, we have no conception of how the mind works. Even the notion works derives from the Mind As Machine metaphor.

(Lakoff and Johnson 1999, page 248)

With those two warnings in place, I will do my best to sort out the troublesome issues and make clear what the questions are and why there is so much confusion. This badly needs doing. Nonspecialists who are interested in "language and the brain" today can buy and read half a dozen fat books, all current, all by qualified experts, only to find that each one says all the others are wrong and offers interminable arguments to that effect. I had a student in a linguistics course once whose major was engineering; a few weeks into the course, tried past endurance, he shouted, "If engineers did engineering the way you linguists do linguistics, every bridge in America would fall down!" He was probably right.

We try hard, I must point out. The subject of linguists' scientific investigation is nothing like chemicals or rocks or beetles, items that can be measured and weighed and poked at. It's not like planets and stars, items that in spite of being enormously inconvenient for human investigation can nevertheless be examined with sufficiently powerful telescopes and probes and the like. Rather, it is located somewhere—or perhaps somewhen, and more probably both—in the human brain and/or mind. If Karl Pribram's model of the brain as a hologram is correct, it is perhaps located everywhere—and everywhen—in the human brain and/or mind. Our access to the brain is very limited; as for access to the mind, we don't know where the mind is, or even whether it makes any sense to talk about it as having a location. Nevertheless, we linguists press on, determined to do our work as science rather than as speculation, and we should get credit for that.

In the context of this chapter, we're looking for answers to many questions; the six below represent what seems to me to be the barest minimum.

1. How is a single native language learned? What goes on in the brain and mind?
2. How does the learning of an additional native language or languages take place? What goes on in the brain and mind?
3. How does the learning of one or more additional languages—not as native tongues but as foreign ones—take place? What goes on in the brain and mind?

For those first three questions, we also want to know whether the answers are different for infants (as opposed to older children and adults) and whether there's a "dose-related" difference that makes learning one language substantially and significantly different from learning four or learning fourteen.

4. What physiological effect(s) and psychological effect(s) do these learning processes have on the brain and mind?
5. Is the answer to #4 different for monolingual versus multilingual persons?
6. Is there some separate language "faculty" or "organ" that is responsible for all this?

These questions are the subject of ongoing research in the cognitive sciences and in their applications in fields like medicine, education, law and law enforcement, and more; now that we ask computers to produce and understand roughly natural speech, they're important in computer science and robotics and engineering. The answers we have at the moment are, in my opinion, only in the most preliminary and tentative stage, although they are sometimes presented in the literature in a very authoritative style.

Much is being written right now about the metaphor and model in which knowledge (loosely speaking) takes the form of connections laid down between and among neurons, with the claim that connections frequently activated become strong and that unused connections disappear. Because this model is biological instead of mechanical, the disappearance in question is literally by death—the death of the unused neurons. One way of thinking about that would be to say that it's as if the newborn brain is a huge assemblage of dots, and learning is a matter of connecting those dots by drawing lines between them; the more you go back and forth over the line you draw, the harder you bear down with your pencil, the less likely it is that that line will fade away or be erased. That's different from imagining switches that are to be turned on or off; on the other hand, you can't help thinking about what happens when you go over a line so many times and with such force that you poke a hole in the paper.

Overview

Linguists and neuroscientists, cognitive scientists of all kinds, are split into factions, each of which considers the others to be not just plain vanilla wrong but stubbornly and perhaps indefensibly wrong. (I won't be discussing all of those factions here; if I tried to do that, I'd inevitably fail, because by the time the book appeared there'd be at least three new ones.)

One major division is between those who believe that human infants are born with some sort of innate language learning "equipment" or "ability" and those who disagree with that idea. That split has an additional branch, growing out of Lakoff and Johnson's claim in Philosophy in the Flesh that innateness is no more relevant in this context than gender and pagination are in the context of God-talk. The innateness versus noninnateness split, they contend on page 507, "makes very little sense given what we have learned about human brains." Another contentious split exists over whether or not language learning—innate, noninnate, or other—is a special and unique kind of learning. New factions and new proposed theories continue to appear; for example, an entire volume has just been published promoting the "emergentist" approach to the subject, which is neither explicitly for nor explicitly against innateness but suggests a compromise. (See MacWhinney 1999.)

My own votes are for the innateness hypothesis and for language learning as unique; I find the evidence from research on sign languages particularly compelling, because it is our first real evidence of children constructing a language from scratch, on their own. (See Aguilera-Hellweg, 1994.) This puts me squarely in one of the factions and will be obvious as you read. I would plead for moderation among the opposing camps, however, since all of us at this point are surely wrong to some degree.

Human infants are born unable to speak or sign or understand any language. However, their brains and minds are in some sense prepared in advance to recognize languages and to work out their rules from raw language data. By "raw data" I mean data that come to the infant in the form of ordinary daily speech and experience rather than as systematic, organized language lessons or instruction. Except in the case of very severe mental retardation or trauma or physical incapacity, children go on to acquire native command of the language or languages they're exposed to in their language environment, whether those are sound languages or sign languages. Children do this in spite of the fact that the data provided to them are often limited or defective, and vary greatly from one environment to another. They go through a uniform series of stages that Steven Pinker has labeled, charmingly, "Syllable Babbling, Gibberish Babbling, One-Word Utterances, Two-Word Strings, and then All Hell Breaks Loose" (Pinker 1994, page 269).

Suppose that from birth you provide your infant with carefully designed daily English lessons taught by a well-qualified native speaker of English. You can then expect that by roughly the age of twelve months the child will begin to use a few words of English, and that by roughly the age of five years the child will speak English as an adult does. The five-year-old may say "wabbit" for "rabbit" and will not have an adult vocabulary or an adult's communication skills—but the child will know all the basics just as well as the adult does. The child will be able to make statements, ask questions, issue commands and threats and promises and verbal attacks, carry on a conversation, argue with other people, indicate past and future time, specify singular and plural, indicate negatives—all the basics. The child won't be able to recite the rules he or she uses in doing all the these things, of course. But adults can't recite those rules, either, unless they're trained linguists—and even then, we don't know how accurate the rules are. As any linguistics student can tell you, the alleged rules, and their forms, keep changing.

It's important not to let yourself be confused by the fact that you, the educated adult, are able to recite all sorts of grammar rules you were taught in English classes. If you try to state the rule you use to do something as simple as forming a yes/no question such as "Did you see a football game today?" you'll discover that you can't do it. As a linguist, my formulation of that rule for practical everyday purposes is "Move the first auxiliary verb into a position immediately to the left of the surface subject of the sentence," which isn't the sort of rule taught in English classes. Nevertheless, and regardless of whether my suggestion for the rule is correct or not, you do know that rule, whatever its exact form may turn out to be, and so does every English-speaking small child. Each time you form a yes/no question accurately, you demonstrate that knowledge, as does the child. I once asked one of my linguistics professors why we don't state the rule as "Move the surface subject of the sentence into a position immediately to the right of the first auxiliary verb in the sentence," and got a startled look, followed by an explanation that was more elegant but boiled down to "We don't do it that way because we've never done it that way." None of this keeps us from proving, by using the rule, that we really do know what it says.

Suppose that from birth you provide your infant with carefully designed daily lessons in both English and Chinese, taught by well-qualified native speakers. Perhaps you offer the two languages on alternate days; perhaps you offer one language in the morning and the other in the afternoon, every day. The child who gets these lessons will have the same experience as the child discussed previously, but in both English and Chinese. There may be a slight delay in beginning to talk, and the process of mastering the basics may be completed a little later, but that's also true of individual children who learn only one language. That's why we use the phrases "roughly twelve months" and "roughly age five."

(The average timetable holds for all human children and for all languages, but there is a considerable amount of individual variation. There are children—entirely normal children who go on to normal language competence and performance—whose first word comes as early as seven months or who don't master the basics until as late as age eight. There are, however, no children whose first words occur at two months or who don't master the basics until age twelve and who still go on to demonstrate normal language abilities.)

The difference between the two kids in terms of their language development is that the child who gets lessons in both Chinese and English will go through a stage in which Chinese and English are slightly mixed up—but that stage is brief and will be over by roughly age five.

However, in both cases—English only, or English and Chinese—if the idea of language lessons for an infant never enters your head and you simply use language around the child in the ordinary way, the same thing will happen. With or without lessons, talking in the language(s) used in the child's environment will begin at roughly twelve months and the basics will be under control by roughly five years. This remains true in an environment where three or more languages are regularly and routinely spoken by the older children and adults, whether you are sufficiently maniacal to provide lessons in three or more languages or you do nothing of the kind. As stated earlier, if there is an upper limit to the number of languages this is possible for, other than that imposed by the number of waking hours in the child's day, we don't know what it is.

These facts are the same all over the world. There are no human languages for which the average timetable is different in any significant way. If we wanted to support the claim often made by English-speaking Americans that Russian or Navajo is inherently harder to learn than French, we would have to accept the idea that infants who learn Russian or Navajo are more intelligent (or in some other way better equipped at birth to learn languages) than infants who learn French. No evidence exists for such a claim.

This is very different from what happens with other skills that we might want to teach to children. With very rare and unpredictable exceptions, children in our society who aren't given lessons won't be able to play the violin, or figure-skate, or play tennis, or do algebra—even if the other people around them do those things every day. And when lessons are provided for those other skills, there will be extreme variation in what happens. Some children will become experts and others will do badly. Some, in spite of major efforts, will never manage to learn. Some will learn quickly and others will learn slowly, even when intelligence and motor dexterity are essentially the same. The various subparts of the skills will be learned in different orders. And so on. This variation is one of the reasons why many linguists insist that whatever it may be that human beings use for learning languages natively—a language "faculty" or a language "organ" or a language "instinct" or a language something else—it is a separate and different thing from whatever it is they use for the kinds of learning that typically take place as a result of structured lessons and training. (Language is of course not the only skill that children learn without formal lessons.)

Suppose that, as some schools of linguists contend, infants are born with no innate mental equipment for acquiring native languages—and nevertheless manage to do it according to the same timetable and in the same fashion and with the same final results, everywhere in the world. This accomplishment then looks almost magical. If children are born with some sort of advance mental preparation, however, native language learning remains an awesome performance, but a scientific rather than magical explanation seems to me to be more possible. The major controversy among those of us who accept the idea that something about language learning is innate is over what that advance mental preparation is like.

I think it's safe to say that newborn children already have available, as part of their mental equipment, a set of specifications for what human languages can contain and what human languages can do. Some of these may be either/or specifications; others are specifications with a range of values from minimum to maximum. Children don't face the task of working out the grammar of the language spoken around them from scratch, therefore. Their task is instead the less difficult job of eliminating from the universal set of possible grammar specifications all those that don't belong in the grammar of the language or languages they're learning.

Consider, for example, the set of basic orderings of subject and verb and object—three items that are present in every human language—in sentences. Although more than one ordering will always be possible, every language has one that it considers more basic than the others. In English, for example, that basic ordering is subject and then verb and then object, as in "The rabbit ate the lettuce," although "The lettuce, the rabbit ate" is possible in the proper context.

For the sake of discussion, let's think of the set of ordering possibilities as if it were six on/off switches. (Understand, please, that that's strictly a metaphor—we have no evidence for a claim that anything in the brain is actually in a physical form that could be recognized as an on/off switch.) We could then say, figuratively speaking, that there would be one Most-Basic-Ordering Switch for "subject, then verb, then object," one for "subject, then object, then verb," and so on.

By roughly age five, the kids will have to have done three things with regard to this aspect of language:

1. Discovered which order is most basic in the language (or languages) they're hearing and observing around them.
2. Strengthened the "on" position for that switch by repeatedly activating the neurons and neuronal connections that are responsible for it.
3. Allowed the other switches to decay (metaphorically speaking, moved them to the "off" position) by failing to activate the neurons and neuronal connections responsible for them.

What the child who is learning two or more languages natively has to do is slightly different. There may be fewer switches in the set of six to be turned off; and it will be necessary, by roughly age five, to know which switches that are turned on correspond to which of the languages learned. Children can do this with ease.

The ability to learn languages in the way outlined above—from raw data, without formal instruction or study—starts to decline as the child grows older, at varying rates depending on the individual. For many (if not most) people, very little of this ability remains by adulthood. It may be that they then do have to learn foreign languages by using the same strategies and abilities and mental equipment that they use for any other kind of learning.

What we see happening in adult foreign language classes bears that out. There is no uniform worldwide pattern in results. Even large groups, of people who go through identical courses in a language show great variation afterward. Some work very hard and never learn much; others seem to "pick up languages" with little effort. Some, no matter how hard they try, always have a strong foreign accent, while others sound much more native. Some learn well enough to get by in a year or two while others may need five or six years of classes to achieve the same level of skill. That is, we see the same sorts of patterns that we see when people take classes in geometry or tennis. There doesn't seem to be any evidence for a universal adult "language acquisition device" that would correspond to the one infants are equipped with.

The model of language learning I've described here makes it possible to understand why adults find it easier to learn languages that are very much like languages they already know. Consider the subject and verb and object ordering in sentences again. The child is born with all six relevant "switches" for that part of the grammar present. If one language acquired in infancy requires Switch 3 to be turned on and another requires Switch 5, learning that requires no extra effort for an infant or very young child. But in an adult whose only native language requires Switch 3 to be on, all the other switches in the set have long since been turned off. When that adult tries to learn a language requiring Switch 3, the learning task is only to conclude, "Oh, I see; it's just like my native language." But an adult whose native language requires only Switch 5 doesn't have that advantage and must memorize the information controlled by Switch 3 as if it had never been encountered before.

This model also explains why it is that the more languages you learn after you reach adulthood, the easier it is to learn yet another one. The more languages you learn, the more your mental arrangements for language learning will resemble the astonishingly effective ones you were born with—metaphorically speaking.

The physical form (using that term very loosely) of the "language acquisition device" in the brain is a mystery, and the scientific papers discussing it are often equally mysterious. We know what the device has to do when the child uses it. It has to create and test hypotheses about the language being learned, and then use the results—with modifications as needed—to build the grammar. The child tries chunks of sound and body language in an effort to communicate with the people in its environment, and concludes from the results that the chunks are or aren't part of the language being learned. It makes perfect sense to say that a child who has tried to say "cookie" and has been rewarded for that by actually getting a cookie is likely to remember the successful chunk of language and try it again. Because that makes sense, there was a time when scholars (called "behaviorists") claimed that language acquisition was simple trial-and-error learning, much like the way research animals will learn to press a lever if you reward them when they do that by accident. However, we can be absolutely certain (for once!) that can't be right. There's no way a system like that could account for the uniformity in language learning by human children.

It used to be suggested that the LAD was like an old-fashioned computer, for which every decision was either entirely right or entirely wrong, and which could never in even the most trivial way go beyond the bare data put into it by human beings. A more recent proposal has been that the LAD is instead like a neural net. A neural net is a computer setup which is able to decide that a particular conclusion is probably somewhat more likely to be correct than some other conclusion and can constantly modify such judgments, not only on the basis of actual data encountered but also by extrapolation from those data.

Lakoff and Johnson propose in Philosophy in the Flesh that the uniformity in language learning can be explained by the uniformity of human infant experience throughout humankind; perhaps that's correct. Perhaps uniformity of experience—being held, being carried, being fed, eating, sleeping, and so on—is sufficient explanation, and would let us dispense with the Language Acquisition Device. It's an interesting hypothesis. It seems to me to imply that instead of universal grammar there is universal culture—a core of human experience that is part of the definition of being human and from which human language is developed by the "embodied mind." If so, given the link between language and culture, I'm not at all sure that anything very different is being proposed.

It may be that all these metaphors, and Pribram's hologram metaphor as well, are wrong. It may be that the brain handles language learning in a way so very special that we have no accurate image with which to describe it yet and no mechanism for talking about it. For the moment, I will continue to assume the innateness hypothesis.

Now, just having the inborn language device, whatever its form and methods, isn't enough to guarantee a native language. We know for sure, without question, that children can't learn a language without regular exposure to that language in use by others who already know it. No normal human being would be cruel enough to test that concept by raising some kids with no exposure to language at all, just to see what would happen. But it has been tested. There have been enough tragic examples of children in very abnormal circumstances—children whose input of language during childhood was drastically limited, sometimes almost nonexistent—to provide such a test. We know that these children may learn to use language to some degree when they're taken from the deprived environment and placed in better circumstances, but they never achieve normal language competency. However, we don't know what the minimum amount of language data input is; kids can manage with an astonishingly small amount.

In the United States (and many other countries and cultures) the typical situation is that the adult who is most responsible for the care of a child during infancy and the toddler years talks to the child in a variety of language usually called "Motherese" (but by no means restricted to mothers). In Motherese the basic vowel sounds of the language are pronounced with an exaggerated clearness, for example. And conversations often include a kind of reinforcement of language that people would never use with older children or adults. Like this:

CHILD: Doggie hungry!
ADULT: Right! The doggie is hungry!
CHILD: Doggie want lunch!
ADULT: Yes, the dog wants its lunch!
CHILD: Mama feed doggie!
ADULT: You think it's time for Mama to feed the dog.
(And so on.)

Much emphasis has been put on the importance of Motherese in the popular press lately. It has even been used—misused—in stories claiming that children in day care facilities will be held back in life because they don't have their mothers doing this sort of thing at them all day long. There are many cultures (and families), however, in which adults rarely speak to infants and small children, and where Motherese would be viewed as an absurd way to deal with kids; nevertheless, children in those environments learn to speak their native languages just like children elsewhere. It's fashionable to insist that learning can't take place without active practice and participation by the learner; it's fashionable, but it's wrong. Much learning can take place passively, with the learner paying very close attention but taking no overt action. In cultures where this is the favored method, people routinely observe a skill in use for long periods of time, and then one day simply begin using it themselves, competently.

In a perfect world, human children from the very first day of their lives would have abundant language input from the adults around them—whether that input came from family members or from other caregivers, either of whom would serve the purpose. Few actions are more likely to help youngsters grow up to be successful and happy adults in our society than providing them with abundant opportunities to practice the skills needed for human language interactions of every kind. This gives kids an extra advantage in today's technological world, where efficient and effective communication of information has become the single most valuable skill. However, we have plenty of evidence that children are perfectly capable of acquiring their native languages without that sort of treatment and attention. They must have exposure to language data; but it doesn't have to be abundant, it doesn't have to include the "Motherese" material, and the data don't even have to be very good.

As stated elsewhere in this book, it was once taken for granted that children who acquired more than one language natively would have somewhat slower language development overall, sometimes coupled with temporary problems related to language in the early school years—but this has been proved to be false. Multilingual children, like any other children, can experience problems at school as a result of all sorts of factors, but there's no cause-and-effect relationship between those problems and multilingualism itself.

We've now come to the end of those things that we can be reasonably certain about, and are moving into areas where we are still only able to speculate. Such as: What happens inside the brain and mind when a native language—or more than one native language—is learned? What happens to the brain and mind? How is that different from what happens when languages are learned later in life, with varying degrees of fluency but not natively? How much do we actually know about such things?

We know very little, and what we know is based on tenuous evidence, often evidence gathered from people whose brains were very different from normal brains due to illness or injury. Reports about brain research appear regularly in the media because brain research is perceived as "sexy." Unfortunately, such reports tend to do little more than increase the general confusion, which is why many scientists behave in a way that gets them accused of being elitists determined to keep their knowledge away from the public.

Well, that's approximately what happened.

Let's look at a recent sequence of research relevant to our discussion in this chapter, and some media reports on that research. Here's a portion of a news release that appeared on the Internet in the summer of 1997 (with emphasis added by me):

In a study that sheds new light on how the brain organizes language, researchers at Memorial Sloan-Kettering Cancer Center report in the July 10 issue of the British journal Nature that the organization of the brain's language-production region in bilingual individuals is directly related to whether they learned a second language as toddlers (simultaneously with their native language) or as young adults … " A second language acquired during the teenage years, which is late in developmental life, is represented in the brain in a separate location from the native language," explains senior author Joy Hirsch, PhD.

The Wall Street Journal ran a story about this study in its July 10, 1997, issue, written by science reporter Ron Winslow and titled "How Language Is Stored in Brain Depends on Age."

"Toddlers who learn a second language along with their native tongue," Winslow wrote, "store this capability in a single sector in the brain, the researchers discovered. But if the second language is acquired later … the brain designates a separate area for processing it." The New York Times, on July 15, ran a story by Sandra Blakeslee headlined "When an Adult Adds a Language, It's One Brain, Two Systems," reporting that "a new study has found that second languages are stored differently in the human brain depending on when they are learned."

The impression that even careful readers would get from the news release and the two newspaper stories is that in the current stage of knowledge about language and the brain we are able to think of the brain as if it were a cupboard, and that we know where languages are "stored" in that cupboard. We can visualize it: Here's the spot on the shelf where little Mary stashed the English and Turkish she learned before the age of three in her multilingual household; here are the two different spots on the shelves where Johnny stashed first the English he learned as an infant and then the Turkish he learned when his family moved to Turkey after he turned sixteen. Scientists can, figuratively speaking, go to those cupboards, open their doors, and poke at the stored languages. The news reports even identify locations for us, referring to "Broca's area" and "Wernicke's area." It's all very impressive, and confidence-inspiring. We know so much! Surely, if we want to know more, we just have to go check it out.

The problem is that the news stories—from a distinguished medical facility and two distinguished newspapers—are misleading. We can reasonably assume that the researchers involved have been tearing their hair out in frustration over these accounts.

Here's what actually happened.

Sloan-Kettering had twelve bilingual patients—six who had learned both languages as toddlers, and six who had learned a second language as teenagers—scheduled for brain surgery. All had to undergo fMRI (functional magnetic resonance imaging) brain scans. The fMRIs, which allow you to see what parts of the brain are active by tracking the flow of blood to various locations, are done to provide the surgeons with individualized brain mapping; this increases the doctors' chances of carrying out surgery without damaging the parts of the brain most responsible for particular functions in that patient. Graduate student Karl Kim, Dr. Hirsch, and their associates asked the twelve patients to talk to themselves silently during the fMRIs about things they had done the day before, first in one of their languages and then in the other. (Talking aloud would have interfered with the fMRI procedure.) The scans were then compared, with particular attention to the two locations in the brain known as Broca's area and Wernicke's area, both of which are known to be involved in language processing.

For all twelve patients, activity in Wernicke's area showed no differences, but there was an interesting difference in the amount of activation in Broca's area. The six patients who learned both languages at the same time in infancy all showed increased activation of just one region in Broca's area. In the six who had learned one language in infancy and the other language in their teens, two separate regions in Broca's area showed activation. In addition, evidence obtained through other Sloan-Kettering patients who weren't involved in this specific study supports the idea that for each additional language learned in the teenage years or adulthood, a separate region in Broca's area would be activated in this same fashion.

That's what happened. Now, what does it tell us? First I need to mention some facts that the news stories didn't bother with or simply trampled into the ground:

1. Twelve research subjects make a very small study. It's typical of brain research that only a handful of subjects are involved, for the obvious reason that getting access to human brains is difficult and expensive. In this study there was the serendipitous fact that the fMRIs had to be done anyway for medical reasons, and the research task was one that could be done at the same time without creating problems. In most areas of human research, studies involving only six to twenty subjects wouldn't be looked at as remotely respectable. An exception is made for brain studies; but brain researchers are well aware, when publishing results based on twenty brains, that if they had been able to get access to twenty-one or twenty-two they might have found something entirely different.
2. Talking to yourself silently, in your mind, is not identical to talking out loud. It's called "covert subvocalization," and we can't be certain that conclusions based on covert subvocalization can be extended to ordinary spoken language.
3. Broca's area is described, in the news stories and elsewhere, as devoted to speech production and grammar. (Wernicke's area is usually described as devoted to interpreting the meaning of language.) But Broca's area—the area where the difference was found—is also deeply involved in controlling motor activities that are part of language. Wendy K.Wilkins and Jennie Wakefield (1995, page 170) write that Broca's area is specialized for "the hierarchical structuring of information in a format consistent with a temporally ordered linear sequence reflective of that structure," which, translated, is the outlining of information. Nothing that we know about Broca's area leads to the conclusion that languages are "stored" there. We don't know with certainty that the idea of language "storage" has any validity. How the media got this "storage" idea I can't say; presumably something one of the researchers said during an interview—in an attempt to explain specialized information to nonspecialists—was misinterpreted, and the misinterpretation was then shared and passed on to the public.

In summary, what we actually did learn as a result of this study is that we have one more piece of evidence for two things that were already well known: (1) that there appears to be some sort of neurological difference in language processing between monolingual persons and multilingual persons; and (2) that there appears to be some sort of difference between language processing in people who become multilingual in infancy or early childhood and those who become multilingual later in life. None of this is news. That doesn't mean that the results of the study aren't important. Evidence in the field of language and the brain is so slim that every additional scrap is precious—but that doesn't make it news. And it's not in any way surprising, given what we believe that we already know. Its primary value is that it supports what we already know.

Language and the Brain

The news release and the newspaper articles caught my attention because I was involved in trying to sort out the answers to my question about whether my multilingual respondents feel like "a different person" when using one or another of their languages. Suppose that when I looked at my database of questionnaires I had found that all those who answered with a yes had learned their additional language(s) as adults, and that all those who anwered with a no had learned their languages simultaneously in very early childhood. Or, somewhat less plausibly, the other way around. That would have been extremely interesting! I would have wanted to pursue the matter. But as is usually the case with such things, it didn't turn out that way. I found no correlation whatsoever between the distribution of yes and no answers and the ages at which my respondents learned their languages.

The model of language learning described in this chapter does answer the "how" questions in the sense that we can describe the process and its outcomes, as demonstrated in the outside world by human behavior, in considerable detail, for all human children. Our answers to the questions about what goes on "in the brain and mind" are still only metaphorical, however.

It does appear to me that language learning before adulthood is a different process from other kinds of learning, although this idea continues to be controversial. If it's correct, it follows that there is a language acquisition device of some kind that's responsible—but only logically, and only metaphorically.

It's possible that every neuron in the brain is involved in language learning and that the process is distributed and duplicated throughout the entire brain; it's possible that our perception of certain areas as being "specialized" for language is a misunderstanding. We have examples of young children who, despite having lost huge portions of their brains to surgery or trauma, develop normal language abilities. At the moment, specialization seems plausible, especially after puberty—but we could be wrong. The brain (and presumably the mind as well) has such an enormous capacity that it doesn't have to conform to our ideas about scientific economy and careful stewardship of resources. We just don't know. Not yet.

When I was studying linguistics in the late 1960s and early 1970s, one of the things my professors talked about was the problem of how meaning is assigned to sentences by the human beings trying to understand them. The profs explained that to a human being it seems that the language-processing part of the brain would have to prevent meaningless sentences from occurring—but that that's just a human prejudice. The brain could just produce any old string of stuff that happened to come along and then throw out the bad ones at the very last minute, in a fashion that scientists would consider disorderly and unscientific. It's like mail order companies that send out every item of merchandise in the same large container. When you get a wristwatch in the mail in a box that would hold a large clock, it strikes you as a waste of resources. But the companies have learned that it's actually cheaper in the long run to throw everything into identical boxes in a single operation than to maintain an inventory of different-sized containers and spend time sorting things out to fit them. It may be that the brain uses principles like that, instead of acting in a way that seems more "brainlike" to us as observers heavily indoctrinated in the prejudices of Western science.

Is there anything we can do about any of this? Now we are truly in uncharted territory. Our past experience with trying to teach languages, or trying to interfere in any other way with the learning and development of languages, indicates that our abilities along these lines aren't impressive. As to whether we should do any of the few things we're able to do—that question is the subject of this book.

It is all too easy to underestimate the power of language, the power of linguistic science, and the power that is represented by mastery of a language. We have an unquestioning respect for the body of knowledge in fields such as physics and astronomy and medicine, and we're awed by people who can routinely make use of such knowledge. But we take the daily use of language for granted. Why? For the very simple reason that everybody, including tiny children, can do it.

If speaking a language were like brain surgery, learned only after many long years of difficult study and practiced only by a handful of remarkable individuals at great expense, we would view it with similar respect and awe. But because almost every human being knows and uses one or more languages, we have let that miracle be trivialized into "only talk." We forget, or are unaware of, the power that language has over our minds and our lives; we use that power ourselves as casually as we use the electric power in our homes, with scarcely a thought given to its potential to help or to harm. We make major decisions about language on the most flimsy and trivial—and often entirely mistaken—grounds. We allow the most basic concepts of language science to remain the cognitive property of a handful of specialists who write and speak about it in a jargon incomprehensible to the rest of us; at the same time, we worry little if at all about the growing ignorance of linguistic science among even high'ly educated people—including those who teach "language courses" in our schools and universities. After all, it's "only talk."

I have tried in this book to show you how unwise and how dangerous such attitudes and practices are, and how important it is for us to refuse to accept them any longer. I have tried to restore some of the sense of wonder that human language is entitled to. I have tried to make it clear that the ability to know and use a language deserves both respect and awe, and that the ability to know and use more than one is even more praiseworthy. I have explained that not only is there no threat to English in the United States from other languages, but it is the other languages of this country that are in danger. I have done my best to point out the implications for humankind of the loss of languages, as well as the possibility that we will find ourselves on this planet with only one language left to us if we're not careful.

I have tried to make a case for the proposition that linguistics should be taught in every school system just as chemistry is; I have offered the daring proposal that it would be more valuable in the curriculum than football. And I have tried to sketch the outlines of some of the unacknowledged motivations behind our current language policies in the United States, especially those that seem completely irrational, such as postponing foreign language courses until after the ability to learn foreign languages has begun to decline.

We go on as we are at our peril.

The End

The Questionnaire

Instructions

Please answer the questions below for me; feel free to use separate sheets of paper if you prefer. E-mail responses are welcome. Questionnaire results are for use in a book that I am writing. Please be sure to let me know if you prefer to remain anonymous—otherwise, you will be given credit for any information that I quote or paraphrase. Feel free to take as much time as you need; I would much rather have your responses later than not at all, and you should not feel rushed. Thank you for your help; if you have questions, don't hesitate to write or call. I'm at 501-559-2273.

Suzette Haden Elgin

Identifying Information, for My Records and for Analyzing the Data

1. Please give me your name, mailing address, e-mail address, and phone number. (Your addresses and phone numbers will be kept confidential.)
2. How old are you?
3. What is your level of education? (Elementary school, high school, college, graduate school, trade school, etc.
4. Please list any degrees or training certificates that you have received.
5. What kind of work do you do?
6. Do you have any special linguistic qualifications? That is, are you a translator or interpreter, a foreign language teacher, a speech therapist, or anything of that kind? Do you have any language disability? Please explain.

Information About Your Languages

1. What is the language that you learned first?
2. What is your second language? How old were you when you learned it?
3. Please list any other languages you know, with the age that each was learned.
4. Which language do you consider to be your native language? (If more than one, please list all of them.)
5. Which language do you consider to be your second language?
6. Please list all languages that you speak comfortably. If you have studied any of them formally—that is, taken classes in them or studied them on your own—please indicate that for me.
7. What languages do you understand easily? (Please indicate any in which you've taken classes or have studied on your own.)
8. What languages do you read easily? (If you've studied them, please indicate that.)
9. What languages do you write easily? (If you've studied them, please indicate that.)
10. Are there other languages that you can "get by" in or have knowledge of? Please list them and explain.
11. When you use any of the languages, do you feel like you are a DIFFERENT PERSON than when you use your native language? If so, could you tell me a little about that?
12. Are there situations in which—even if everyone present knows both or all of the languages—you prefer to use one language rather than another? If so, please explain.
13. Are there situations in which—even if everyone present knows both or all of the languages—you prefer NOT to use a particular language? If so, please explain.
14. Which language or languages do you use at home?
15. Which language or languages do you use at work?
16. Which language or languages do you use socially?
17. Is there a language you use for other purposes? Please explain.
18. What language or languages would you like for your children to learn? Why? (If you don't have children, please answer as you would if you did.)
19. Is there a language you'd rather your children did NOT learn? Why? (If you don't have children, please answer as you would if you did have them.)
20. Are there any languages that you once knew, but have now forgotten? Can you tell me a little about that?
21. Is there anything else you'd like to add about your languages, your feelings about languages, or any related subject?

NOTE: The two questions that follow might be time-consuming for you to answer, and I don't want to impose on you. If you'd rather not bother with them, I will understand. If you can help me with them, however, I would find the information extremely valuable. If you want to choose just one or two of the items instead of dealing with all of them, that's fine. If examples occur to you that have not occurred to me, I would be most grateful to have them. Once I have seen the data I will probably need to come back to you individually and ask a follow-up question or two; I am interested in how bilinguals/multilinguals deal with possible conflict (cognitive dissonance) when their two or more languages express these concepts in drastically different ways. For any that you do answer, I'd be grateful for a literal translation of the items in languages other than English, so that I know what I'm dealing with. If you prefer to send this sort of material by snail mail, I am at PO Box 1137, Huntsville, Arkansas 72740-1137, USA.

23. I am especially interested in metaphorical differences between languages for a number of key areas, both when expressed overtly as metaphor (like English "to pass away" for "to die") and when the metaphor is the literal meaning of the word (as with English "breakfast" for the first meal after the night's fasting). For this purpose, I would be very interested in knowing what words and/or phrases your languages (other than English) use for any or all of the following: to have (or get) an abortion; to die; to be born; to get sick; to be well/healthy; to make a piece of art (or create art); to set up a government; to sin; to get angry; to fall in love; the law of gravity; energy; science.
24. I am also exploring the differences in what are called "unifying metaphors" in bilingual/multilingual speakers and contexts. For example, American English speakers can get almost any project under way by saying, forcefully, "Wagons—HO!" which brings up in consciousness the metaphor of wagon trains setting off across the plains, etc. I would like two pieces of information. First, can you identify a comparable unifying metaphor in your other language or languages? Second, do you feel that any one of the set of such metaphors you use is more powerful and compelling than the other(s)?*

Thank you.

Suzette Haden Elgin

PS: If you have suggestions for other persons who might be willing to participate in this project, I'd be delighted to have their responses as well. And I am of course interested in any comments or criticisms you may have.

*With regard to this question, a generational split exists. People of my own generation (roughly sixty or older) answered the question without hesitation. Younger respondents, however, reported that "Wagons—HO!" was not a unifying metaphor for them; many said they'd never heard it.

The Respondents

NOTE: The languages listed below are those in which the respondent has indicated a perception of fluency—from native fluency at one extreme to the ability to carry on conversations comfortably at the other. Many respondents listed a number of other languages (sometimes many other languages) in which they were literate or could "function as a tourist" or had a variety of other competencies, but which they did not consider themselves able to speak and understand comfortably.

It appeared to me in a number of cases that the respondent was being overly modest about his or her multilingual competencies. In such cases, I have nevertheless reported the response as it was given. (The only exception is that I have added English after making certain that it should be added. Some respondents, though obviously at ease with English, did not list it as one of their languages.) It goes without saying that the respondents' perceptions of their own fluency must be respected. But it's important to know that there are respondents who reported living and working for long periods of time, using a particular language daily for everyday affairs, and who still did not include that language as one they spoke and understood comfortably. This is a clear demonstration of the confusion about what it means to "know a language," as discussed in the Introduction. In my opinion, the multilingual competencies of many respondents are understated below.

I would like to express my deep gratitude to the many respondents who provided me with lengthy responses to Questions 22 and 23, and to those who sent me supplementary pages expanding on their answers to other questions; I know this represented a substantial expenditure of time and energy on their parts:

1. Abichaker, Toufic; Arabic, French, English, Spanish, Italian, German
2. Anderson, Myrdene; English, Norwegian, Saami
3. Andruski, Jean; English, French, German, Mandarin Chinese, Russian
4. Anonymous; Sesotho, English, Zulu, Afrikaans
5. Anonymous; eight fluent languages—requested that I not list them because the distinctive nature of the set would betray identity
6. Anonymous; English, Mandarin Chinese, Cantonese
7. Anonymous; Latvian, English, French
8. Anonymous; English, Spanish
9. Anonymous; Spanish, English
10. Anonymous; German, French, English
11. Anzali, Soheila; Azerbaijani, Farsi, German, English, Turkish
12. Arouca, Karen; Portuguese, English, Spanish, Italian
13. Aylward, Renata; English, Spanish
14. Bauer, Laurie; English, French, Danish, German
15. Bemore, Caroline; Navajo, English
16. Blalock, Lucy Parks; Lenape, English
17. Boihang, Lesedi; Setswana, Xhosa, Zulu, English, Afrikaans
18. Boon, Glendia M.; American Sign Language, English
19. Bregman, L. D.; English, Afrikaans
20. Brewer, Keri; English, American Sign Language
21. Brodine, Ruey; English, Italian
22. Brown, Gordon; English, French, Swedish, Spanish, German
23. Browne, E. Wayles; English, Serbo-Croatian, Russian, Polish, Slovenian
24. Combon, Marlis; German, Spanish, Italian, English, French
25. Chiarenza, Anna Caterina; Italian, German, English
26. Chomphosy, Bounlieng; Lao, Thai, English
27. Chong, Steve; Cantonese, English, Malay
28. Chung, Karen Steffen; English, German, Spanish, Mandarin Chinese
29. Cook, Diana; English, French, Spanish
30. Cyr, Danielle E.; Quebec French, English, Swedish, Italian, German
31. Mark de Clark; English, Afrikaans, Zulu
32. Derzhanski, Ivan A.; Bulgarian, Russian, English, Hungarian, German, French, Italian, Spanish
33. Dorian, Nancy Currier; English, German, French, Scottish Gaelic (East Sutherland variety)
34. Doty, Margaret Ann; English, Italian, German
35. Eck, Karen; English, French
36. Eikamp, Rhonda; English, German
37. Eldredge, Linda G.; English; Signed English for the Deaf; American Sign Language
38. Eroz, Betil; Turkish, English
39. Estival, Dominique; French, English
40. Faris, Michael A.; English, Polish, Polish Sign Language
41. Fidelholtz, James L; English, Spanish, German, Polish
42. Frishberg, Nancy; English, American Sign Language
43. Furiye, Atsushi; Japanese, English
44. Gladstone, Kate; English, Hebrew, French, German, Russian
45. Gluckman, T. J.; English, German, French
46. Gonzalez, Luis Alberto; Spanish, English
47. Gregory, George Ann; English, Spanish, Choctaw
48. Gu, Peter Yongqi; Mandarin Chinese, English, Hebei Chinese
49. Halford, Richard; English, Afrikaans
50. Haraghey, Marjolijn; Dutch, English, German, Afrikaans, French
51. Hardman, M.J.; English, Spanish, Jaqaru, Aymara
52. Herrarte, Caroline; English, Spanish
53. Herzfeld, Anita; Spanish, English, German
54. Hilgendorf, Suzanne K.; English, German
55. Iturrizaga, Dimas Bautista; Jaqaru, Spanish, English
56. Joergensen, Mikala; Danish, English, German, French
57. Julian, Jane; English, Hebrew
58. Kaspy, Lior; Hebrew, English, Arabic
59. Kernberger, Carolyn; English, Spanish
60. Khoza, Zangie; Tsonga, Swazi, English
61. Kirchner, James; English, Czech, German
62. Kondo, Chikako; Japanese, English
63. Kronenberg, Kenneth; English, German
64. Langent, Moeketsi; Sesotho, English, Afrikaans
65. Lattey, Elsa; German, English, Spanish
66. Lenstrup, Marie; Danish, English, German, Swedish, Norwegian
67. Li, Wen-Chao; Mandarin Chinese, English
68. Llanes, Jose R.; Spanish/English
69. Luk, Ching Man; Cantonese, English
70. Maduell, Mariana; English, Spanish, Italian
71. Maesfranckx, Patricia; Dutch, English, German, Swedish, French
72. Mahdi, Waruno; Indonesian, English, German, Russian, Dutch
73. Mamathuba, Elelwani Charity; Venda, English, Afrikaans'
74. Martinez, Alejandro G.; Spanish, English
75. Mclntire, Marina L.; English, American Sign Language
76. Meynard, Yves; French, English
77. Mohamed, Rahima; English, Afrikaans
78. Morrison, Dave; English, American Sign Language
79. Murphy, Linda; English, French
80. O Se, Gavin; English, German, French, Irish
81. Palmer, Chris; English, French, German ,
82. Parkman, Victoria; Russian, Ukrainian, English
83. Pateris, Lehtla; Latvian, Russian, German, English
84. Pomerantz, Ori; Hebrew, English
85. Rasmussen, Allysse Suzanne; English, Spanish, American Sign Language
86. Rebsch, Debby; German, English
87. Reighard, John; English, French, Italian, German, Portuguese, Spanish
88. Resende, Rosana; Portuguese, Spanish, English
89. Robertson, Michael; English, Italian
90. Rose, Alex; English, French
91. Schaufele, Steven; English, French, German, Italian
92. Schober, George; English, German
93. Seek, Maguette; Wolof, French, English
94. Siley, Delphine; French, English
95. Sosa, Juan Manuel; Spanish, French, English, Portugeuse
96. Sotello, Susana; Spanish, English
97. Stallings, Fran; English, Japanese
98. Stenson, Nancy; English, Irish, French
99. Swan, Pamela; Japanese, English, Spanish
100. Troen, Carol R.; English, Hebrew, French
101. Trudel, Jean-Louis; French, English
102. True, Sean; English, Spanish
103. Umans, Meg; English, Spanish, German, French
104. van der Goot, Auke; Frisian, Dutch, English, French, German
105. van Wyk, Karin; English, Afrikaans
106. Verspoor, Cornelia M.; English, Dutch, French
107. Vlijm, Chantal; Dutch, English, French, German, Spanish
108. Vonarburg, Elisabeth; French, English
109. Walling, Espy Rodriguez; Spanish, English, French
110. Whiteley, Colin; English, Spanish, French, German, Catalan, Portuguese, Italian
111. Wilson, Mark A.; English, German
112. Yamamoto, Masayo; Japanese, English
113. Yanjin, Yao; Chinese (Mandarin and Shanghai), English
114. Zeviar, Churyl; English, French, American Sign Language
115. Zoeppritz, Magdalena; German, English, French

The Languages of Respondents

Books

Aitchison, J. 1998. The Articulate Mammal: An Introduction to Psycholinguistics, 4th ed. New York: Routledge.

Asher, R. E., ed. 1994. The Encyclopedia of Language and Linguistics. Oxford: Pergamon Press.

Auer, P., ed. 1998. Code-Switching in Conversation: Language, Interaction, and Identity. New York: Routledge.

Bain, B. 1996. Pathways to the Peak of Mount Piaget and Vygotsky: Speaking and Cognizing Monolingually and Bilingually. Rome: Bulzoni Editore.

Beinfield, H., and E. Korngold. 1991. Between Heaven And Earth: A Guide to Chinese Medicine. New York: Ballantine Books.

Berube, M. 1996. Life as We Know It: A Father, a Family, and an Exceptional Child. New York: Pantheon Books.

Bodmer, F. 1944. The Loom of Language. (Edited by L. Hogben.) New York: W. W. Norton.

Bolinger, D. 1972. Intonation. Harmondsworth: Penguin Books.

Cantoni, G., ed. 1996. Stabilizing Indigenous Languages. Flagstaff: Northern Arizona University. [A Center for Excellence in Education monograph.]

Carson, D. A. 1998. The Inclusive-Language Debate: A Plea for Realism. Grand Rapids, MI: Baker Books.

Cherry, K. 1987. Womansword: What Japanese Words Say About Women. New York: Kodansha America.

Claiborne, R. 1983. Our Marvelous Native Tongue: The Life and Times of the English Language. New York: Times Books.

Cook, V., and M. Newson. 1996. Chomsky's Universal Grammar, 2nd ed. London: Blackwell Publishers.

Crawford, J. 1992. Hold Your Tongue: Bilingualism and the Politics of English Only. Reading, MA: Addison-Wesley.

Crystal, D. 1992. An Encyclopedic Dictionary of Language and Languages. London: Penguin Books.

_____. 1994. Dictionary of Language and Languages. New York: Penguin.

_____. 1998. The Cambridge Encyclopedia of Language, 2nd ed. Cambridge, UK: Cambridge University Press.

_____. 1998. English as a Global Language. Cambridge, UK: Cambridge University Press.

Edwards, J. 1995. Multilingualism. London and New York: Routledge.

Ekman, P. 1985. Telling Lies: Clues to Deceit in the Marketplace, Politics, and Marriage. New York: W. W. Norton.

Elgin, S. H. 1984. Native Tongue. New York: DAW Books. (And forthcoming from Feminist Press, New York, in October 2000.)

_____. 1987. Native Tongue II: The Judas Rose. New York: DAW Books.

_____. 1988. A First Grammar and Dictionary of Laadan. Madison, WI: Society for the Furtherance of Fantasy and Science Fiction (SF3).

_____. 1993. Genderspeak: Men, Women, and the Gentle Art of Verbal Self-Defense. New York: Wiley.

_____. 1994. Native Tongue III: Earthsong. New York: DAW Books.

_____. 1996. How to Disagree Without Being Disagreeable. New York: Wiley.

_____. 1997. How to Turn the Other Cheek and Still Survive in Today's World. Nashville, TN: Thomas Nelson.

Ferre, F. 1961. Language, Logic, and God. New York: Harper and Row.

Festinger, L. 1957. A Theory of Cognitive Dissonance. Evanston, IL: Row, Peterson.

Frantzis, B. K. 1993. Opening the Energy Gates of the Body. Berkeley, CA: North Atlantic Books.

Friedrich, P. 1986. The Language Parallax: Linguistic Relativism and Poetic Indeterminacy. Austin: University of Texas Press.

Gannon, M. J. 1994. Understanding Global Cultures: Metaphorical Journeys Through 17 Countries. Thousand Oaks, CA: Sage.

Gill, S. and I. Sullivan, 1992. Dictionary of Native American Mythology. New York: Oup.

Grimes, B. F. 1992. Ethnologue: Languages of the World, 12th ed. Dallas, TX: Summer Institute of Linguistics.

Grosjean, F. 1982. Life with Two Languages: An Introduction to Bilingual-ism. Cambridge, MA: Harvard University Press.

Hagege, C. 1996. L'enfant aux deux langues. Paris: Editions Odile Jacob.

Hakuta, K. 1986. Mirror of Language: The Debate on Bilingualism. New York: Basic Books.

Hall, E. T. 1959. The Silent Language. New York: Doubleday/Anchor.

_____. 1969. The Hidden Dimension. New York: Doubleday/Anchor.

_____. 1977. Beyond Culture. New York: Doubleday/Anchor.

_____. 1984. The Dance of Life: The Other Dimension of Time. New York: Doubleday/Anchor.

[Index omitted]

The Gentle Art of Verbal Self-Defense

More on the Gentle Art of Verbal Self-Defense

The Last Word on the Gentle Art of Verbal Self-Defense

Genderspeak

Success with the Gentle Art of Verbal Self-Defense

The Grandmother Principles

You Can't SAY That to Me!

Try to Feel It My Way

How to Disagree Without Being Disagreeable

PERSEUS PUBLISHING

Cambridge, Massachusetts

Version 1.0

Many of the designations used by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in this book and Perseus Publishing was aware of a trademark claim, the designations have been printed in initial capital letters.

CIP information for this book is available from the Library of Congress

ISBN: 0-7382-0428-5

Copyright © 2000 by Suzette Haden Elgin, Ph.D.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Printed in the United States of America.

Perseus Publishing is a member of the Perseus Books Group

Text design by Jeff Williams

1 2 3 4 5 6 7 8 9 10—03 02 01 00

First paperback printing, December 2000

Perseus Publishing books are available at special discounts for bulk purchases in the U.S. by corporations, institutions, and other organizations. For more information, please contact the Special Markets Department at HarperCollins Publishers, 10 East 53rd Street, New York, NY 10022, or call 1-212-207-7528.