Any day now Lezlie Cowley Allison, a preppy 25-year-old with a mild Texas drawl, will be packing for a remote corner of the planet, probably somewhere in West Africa. Having recently completed a master's degree in linguistics, she is about to begin the decades of field work necessary to master one of the world's dying oral languages.

At the beginning of her studies, when she must find a way of communicating with the villagers among whom she will be living, she will very likely have only one Western companion - her husband, Sean Allison, who shares her interests. She expects to give birth to the couple's first child in November. Allison says she is not particularly anxious about the health aspects of raising an infant in rural Africa, although she admits, "I guess it would be better not to have a dirt floor."Allison's primary motivation is not academic but spiritual. She is a missionary-in-training, soon to become a full member of Wycliffe Bible Translators, a Protestant organization based in Huntington Beach, Calif., that produces more missionary linguists than any other group. These linguists are now translating the New Testament and parts of the Old Testament into some 1,200 languages around the world.

"After 10 years," Allison says, "I know I'll be able to say that I've analyzed a grammar and begun to help a group of people understand that God speaks directly to them."

The fact that there are so many Bible translators and so many translation projects under way does not mean that a vast number of people will be newly exposed to Christian ideas, because portions of the scriptures already are available in the native tongues of about 97 percent of the world's population. Instead the projects are noteworthy because of the types of languages involved: many of them are spoken by fewer than 10,000 people, do not exist in written form, and have never been studied by outsiders.

The missionaries who take on the enormous task of producing scripture in these languages become crack linguists. They are the largest army of field linguists in the world, equipped with state-of-the-art linguistic software running on battery or solar-powered laptop computers. And although the translators themselves earn only a pittance, funding for these projects tops $100 million a year.

The quantity and quality of the missionaries' linguistic work has not gone unnoticed in the secular world. Over the past 50 years a symbiotic relationship has evolved between many Bible translators and university linguists - a latter-day version of the earlier fruitful collaboration between missionaries abroad and naturalists back home.

Today's missionaries contribute a gold mine of painstakingly gathered linguistic data, which helps scholars better understand the universals and idiosyncrasies of language. They also are helping to preserve a record ofminority languages at a time when such languages are rapidly disappearing, victims of many of the same forces that have brought floral and faunal species to extinction.

Wycliffe Bible Translators, named after John Wycliffe, a 14th-century religious figure who initiated the first English translation of the complete Bible, specializes in the most precarious languages. Approximately 20 other organizations around the world, including the American Bible Society, the United Bible Societies, and New Tribes Mission, sponsor Bible-translation efforts, but Wycliffe's linguists are pre-eminent in working with the oral languages of remote groups.

Most of Wycliffe's linguists, including Lezlie Cowley Allison, have been initially trained by the group's sister organization, the Summer Institute of Linguistics, which offers rigorous college- and university-affiliated classes in Dallas, where its headquarters are located, and in Oregon, North Dakota and eight other countries. (Four of these sites now have year-round programs, so SIL's name has become a misnomer.)

Since SIL was founded in the 1930s, its members have produced linguistic data on about 900 languages and have completed 400 New Testament translations. Currently more than 6,000 SIL translators, literacy specialists and support-staffers are working on six continents. (It should be noted that the names Wycliffe and SIL are often used interchangeably, even by the organizations themselves, although Wycliffe is generally the name that fund-raisers use in local Protestant churches, and SIL is used in academic circles and abroad. For the purposes of this article, I'll refer to both as SIL.)

SIL is not without its critics. Over the years the organization has faced charges that it has cooperated with the CIA, organized counterinsurgency camps in Latin America, and trafficked in drugs and gemstones. Such allegations, which have never been substantiated but have led to SIL's expulsion from Ecuador, often surface when the missionaries' presence cuts into the access that traders or businessmen have to indigenous people or their land.

A more credible, and therefore more damaging, line of attack comes from secular anthropologists, who find SIL guilty of all the traditional missionary vices: paternalism, ethnocentrism and cultural imperialism. But even though these critics object to SIL's spiritual ends and means, many of them willingly concede that the organization's linguistic work is unimpeachable.

Twenty-five years ago, when only a handful of universities had their own linguistics departments, SIL attracted many prospective field linguists who had no interest in Bible translation. Today most of the 190 students at SIL's Dallas headquarters are, like Lezlie Allison, Protestant evangelical Christians. Most are opposed to abortion and skeptical about assumptions of evolutionary theory.

But despite the religious uniformity, signs of religious-secular cross-pollination are abundant. Many SIL faculty members hold Ph.D.s from secular institutions such as the University of Michigan, the University of Pennsylvania and Cornell University. Courses draw on material from secular theoreticians teaching at the Massachusetts Institute of Technology, the University of Chicago and Yale University.

For its part, SIL has prepared an academic-mission statement that ranks "contribution to the academic world" second only to "quality of our work" in its list of priorities. And a 600-page bibliography lists some 14,000 scholarly books and articles written by SIL's field members since the 1930s.

"All of us have Christian motivations - that's what allows us to stick it out in unpleasant situations," says George Huttar, the head of SIL's academic programs, who specializes in Caribbean Creole languages. "But that doesn't mean we're in it just to produce Bibles. We feel an ethical responsibility to gather quality linguistic data and make it available, especially in cases in which we're the only members of the academic community to have access to a language."

The quality of the missionaries' work, both academic and religious, depends heavily on their receiving regular feedback from secular peers. Scholarly endeavors are also an effective rejoinder to those who are skeptical of work produced by evangelicals.

"Quite frankly, some missionary organizations are still boldly leading their recruits into the 1890s," says Thomas Headland, an SIL anthropologist who has spent 25 years working with the Agta hunter-gatherers, of the Philippines. "We have to prove ourselves through the quality and usefulness of our linguistic work."

Bible translators have a long history of linguistic acumen. The writing systems of several Indo-European languages, for example, were first devised by missionaries during the course of translation. The missionary Ulfilas is generally given credit for putting Gothic into writing in the fourth century, Mesrob for Armenian in the fifth century, and Methodius and Cyril (as in "Cyrillic") for Slavic in the ninth century.

The guides to several Latin American Indian languages that Jesuits produced in the mid-1500s are still used by researchers today. But linguistic work was historically a fortuitous side effect of translation - not a primary focus, as it is today.

The transition can be traced to the 1940s and 1950s. This was the era when new scientific methods and theories at once opened a world of previously unstudied languages to translators and created a booming market for data from those languages. Many missionary linguists were swept into the academic mainstream.

The man who, surely, has done the most to blur the distinction between Bible translators and academic linguists is Kenneth Pike. Pike began his career with SIL in the 1930s as a Bible translator among a group of Mixtec Indians in Mexico, periodically returning to study at the University of Michigan under the tutelage of such renowned linguists as Edward Sapir and Charles Fries.

In 1948 he joined Michigan's faculty, securing a unique arrangement that allowed him also to serve as SIL's president and to visit hundreds of SIL translators in the field. During the next 30 years Pike published several books on linguistic theory that became classics, served as president of the Linguistic Society of America, accepted a chair in Michigan's linguistics department, was elected to the National Academy of Sciences, and received honorary doctorates from the University of Chicago and the Sorbonne.

Now 83, Pike lives year-round in Dallas, but he is not nearly ready to retire. (He gave up water polo only in his late 60s.) He says he has achieved many of his personal ambitions during his career, high among them "to make the scientists stand up and take notice."

Two key goals nag at him: perhaps as many as a thousand languages still need to be cracked and translated, and he still longs to find and inspire one or two young scholars who will touch off the next revolution in linguistic theory.

"There's much more work to be done," he says. "We've gotten to be very good at recruiting translators and turning them into linguists. But only a few of us have focused on linguistic theory."

Even if one of Pike's disciples does become the next Noam Chomsky, SIL's claim to fame will probably continue to be not theory but grassroots field research, which now falls under the rubric "descriptive linguistics." It is painstaking work, and in the case of many minority languages, linguists must go in cold, without the benefit of a lingua franca such as Spanish, French or Portuguese to serve as a mediating tongue.

A prerequisite to all fieldwork is establishing a place within the community, which usually takes many months of silent observation and the patient guidance of colleagues who have worked with related minority groups.

The goal is for the researcher, once he or she is able to begin communicating orally, to assume a social position that allows access to all members of the minority group, from leaders to paupers. That accomplished, he or she begins a long period of listening to and recording the speech of native speakers to determine which basic sounds make up the language.

The building blocks of language are called phonemes, and they are a finite set - a set defined by the possible contortions of the vocal cords, mouth, lips and tongue. (English-speakers use about 44 distinct phonemes; Marghi, a language of Nigeria, uses at least 70.) To transcribe the phonemes of a language and begin to establish a written form of it, researchers use or borrow heavily from the standardized International Phonetic Alphabet, which is based on Roman characters but spiced with diacritical marks and Greek letters.

Many languages in Africa, Asia and Latin America require an additional system of notation to account for pitch levels. In these languages a word can have as many as five different meanings depending on the mixture of tones used to pronounce it. (Intonation is important in all languages, but in many it is used to change emphasis rather than denotation: "This is MY son" vs. "This is my SON.")

To help analyze tones - one of the trickiest tasks in descriptive linguistics - researchers sometimes ask native speakers to whistle their words. (In a few African cultures, however, whistling is taboo.) Another technique is to "play" words on simple flutes or on glasses filled with water to different levels, one for each register of the tone language. Nowadays field researchers also have the benefit of computerized tone analyzers, which convert live or recorded speech into graphic, EKG-like displays.

Once an alphabet has been established, researchers must transcribe words and then stories, such as historical or mythical accounts. This is the raw data enabling linguists to understand how sounds are strung together to form units of meaning: words, phrases, sentences, passages.

Despite the irregularities inherent in all languages, linguists work under the assumption that word structure and grammar can be formalized into a set of rules that re-create in methodical, adult terms the formulas every native speaker internalizes as a child. Finding those rules requires a mind that thrives on puzzles. It is no coincidence that many of SIL's most prolific scholars came to linguistics with a background in mathematics, engineering or computer science.

The chief end products of the descriptive linguist's work are an analysis of the sound system, a usable alphabet, a grammar and a dictionary. SIL has always published these guides, but other kinds of valuable field data and analyses often gather dust - literally in old shoe boxes.

Computers have begun to change that. SIL is a world leader in the creation of software for field linguists, and it will soon complete a series of linked computer programs to accommodate (and standardize) all linguistic data, including digitized sound recordings.

After a linguist types in a particular story, for example, the computer can scan the new document for all the words and constructions in it that the linguist has previously encountered. The computer then automatically annotates the text it recognizes with various levels of linguistic description, such as part of speech and English translation. The researcher can then touch up the annotation by teaching the computer new rules and words for the next time around.

When the researcher is translating a verse from the Bible, he or she can also call up on the computer screen many different English translations of the verse, plus the Greek or Hebrew original; search a 200,000-page library for exegetical commentary; refer to the same passage in an obscure but related tongue; and scan all the stories the computer has annotated for various ways of expressing a concept.

The series of linked computer programs, known as CELLAR (Computing Environment for Linguistic, Literary and Anthropological Research), is expected to revolutionize field research.

"The average university linguist is on his own when it comes to software," says Gary Simons, SIL's chief computer expert, who earned a Ph.D. in linguistics from Cornell for his work on several languages from Papua New Guinea and the Solomon Islands. "He doesn't have this kind of support from his institution. We see our own members as a fraction of our software audience."

SIL will sell the software to outside researchers. It charges buyers only the cost of printing the documentation that comes with it and the cost of several blank disks.

Descriptive data, especially when computerized, is always in demand in academia. "Theoreticians have long tested their hypotheses about language using only German and English," says Eugene Loos, a former international linguistics coordinator for SIL and an expert in several Peruvian Indian tongues. "Now they want to test them on other live languages from various parts of the world. That's the only way to find real universals."

To that end, Oxford University, which is interested in increasing its holdings of non-Indo-European linguistic data, would like to expand by adding much of SIL's newly computerized information to its system. The University of California at Berkeley has keyed in many of SIL's published dictionaries of Amazonian languages.

SIL linguists have several favorite stories of how missionary field data has debunked existing theories. For example, Geoffrey Pullum, a noted English syntactician, argued in a 1977 article that languages in which objects most commonly precede both subjects and verbs do not exist, because the human mind would not allow it.

One of Pullum's students, a soft-spoken SIL missionary named Desmond Derbyshire, spoke up in class to explain that he was a fluent speaker of such a language, Hixkaryana, from the Brazilian Amazon.

"Pullum was skeptical for about a week while we looked through my data," says Derbyshire, who is now in Dallas. "But I was right. Object-initial is the most natural form of expression among these people. For example, a Hixkaryana speaker says, `Toto yoskano kamara' - `Man killed jaguar' - to express the English sentence `The jaguar killed the man.' "

The Summer Institute of Linguistics has received several grants from development agencies, including the United Nations Educational, Scientific and Cultural Organization, to promote minority-language literacy programs. These organizations seek to preserve a record of minority languages, but they have an economic-development agenda as well: indigenous peoples cannot be integrated into a larger national economy if they do not know the national language, and, it is believed, they tend to be more successful in learning it if they first become literate in their mother tongue. Anthropologists, it should be noted, have mixed feelings about the value of speeding minority peoples into the economic mainstream.

Still, the future of Bible translation is linguistic "technology transfer" from outside organizations like SIL to homegrown groups. Because the level of general education has improved throughout the Third World, SIL is drawing more international students who want to be trained to translate scripture into their own and other minority languages.

Lezlie Allison's generation may be the last in which native English-speakers lead translation efforts. United Bible Societies, for example, no longer sends Western linguists into the field, and instead provides full-time regional consultants who supervise local recruits.

The new approach will not mollify angry anthropologists, who take issue with the religious messages but not necessarily with the nationality of the messengers. It will, however, assure a steady supply of quality field linguists at a time when minority languages are disappearing faster than they can possibly be studied by outsiders.

Whatever the nationality, the impetus for the work will remain something that is rare in this world: a common vision. "I dream of seeing all the people of the world joining together in a grand symphony in praise of God," says Robert Longacre, one of SIL's foremost theoreticians. "Each group would play its own instrument - its own language - whether it be the little piccolo or the bass drum. And all would be reading from the same score."