Über die Autorin bzw. den Autor

Mark C. Baker is a professor in the Department of Linguistics and the Center for Cognitive Science at Rutgers University. He lives in Camden, New Jersey.

Auszug. © Genehmigter Nachdruck. Alle Rechte vorbehalten.

The Atoms of Language

Basic Books

Chapter One

The Code Talker Paradox

Deep mysteries of language are illustrated by an incident thatoccurred in 1943, when the Japanese military was firmly entrenchedaround the Bismarck Archipelago. American pilots had nicknamedthe harbor of Rabaul "Dead End" because so many of themwere shot down by antiaircraft guns placed in the surrounding hills. Itbecame apparent that the Japanese could easily decode Allied messagesand thus were forewarned about the time and place of each attack.

    The Marine Corps responded by calling in one of their most effectivesecret weapons: eleven Navajo Indians. These were membersof the famous Code Talkers, whose native language was the one cipherthe Japanese cryptographers were never able to break. TheNavajos quickly provided secure communications, and the area wassoon taken with minimal further losses. Such incidents were repeatedthroughout the Pacific theater in World War II. Years after the end ofthe war, a U.S. president commended the Navajo Code Talkers withthe following words: "Their resourcefulness, tenacity, integrity andcourage saved the lives of countless men and women and sped the realizationof peace for war-torn lands." But it was not only their resourcefulness,tenacity, integrity, and courage that made possibletheir remarkable contribution: It was also their language.

    This incident vividly illustrates the fundamental puzzle of linguistics.On the one hand, Navajo must be extremely different from English(and Japanese), or the men listening to the Code Talkers? transmissionswould eventually have been able to figure out what theywere saying. On the other hand, Navajo must be extremely similarto English (and Japanese), or the Code Talkers could not have transmittedwith precision the messages formulated by their English-speakingcommanders. Navajo was effective as a code because it hadboth of these properties. But this seems like a contradiction: Howcan two languages be simultaneously so similar and so different?This paradox has beset the comparative study of human languagesfor centuries. Linguists are beginning to understand how the paradoxcan be dissolved, making it possible for the first time to chart outprecisely the ways in which human languages can differ from one anotherand the ways in which they are all the same.

* * *

Let us first consider more carefully the evidence that languages canbe radically different. The Japanese readily solved the various artificialcodes dreamed up by Allied cryptographers. Translating a messagefrom English to Navajo evidently involves transforming it inways that are more far-reaching than could be imagined by the mostclever engineers or mathematicians of that era. This seems more remarkableif one knows something about the codes in use in WorldWar II, which were markedly more sophisticated than any used beforethat time. In this respect, an ordinary human language goes farbeyond the bounds of what can reasonably be called a code. If thedifferences between Navajo and English were only a matter of replacingwords like man with Navajo-sounding vocabulary likehastiin, or putting the words in a slightly different order, decodingNavajo would not have been so difficult. It is clear that the characteristicsone might expect to see emphasized in the first few pages ofa grammar book barely scratch the surface of the complexity of atruly foreign language.

    Other signs of the complexity and diversity of human languagesare closer to our everyday experience. Consider, for example, yourpersonal computer. It is vastly smaller and more powerful than anythingthe inventors of the computer imagined back in the 1950s.Nevertheless, it falls far short of the early computer scientists? expectationsin its ability to speak English. Since the beginning of thecomputer age, founders of artificial intelligence such as Alan Turingand Marvin Minsky have foreseen a time in which people and computerswould interact in a natural human language, just as two peoplemight talk to each other on a telephone. This expectation wascommunicated vividly to the world at large through the 1968 movie2001: A Space Odyssey, in which the computer HAL understood andspoke grammatically perfect (if somewhat condescending) English.Indeed, natural language was not even considered one of the "hard"problems of computer engineering in the 1960s; the academic leadersthought that it would more or less take care of itself once peoplegot around to it. Thirty-five years and billions of research dollarslater, their confidence has proved unwarranted. It is now 2001, andthough HAL's switches and indicator lights look hopelessly out-of-date,his language skills are still in the indefinite future. Progress isbeing made: We only recently achieved the pleasure of listening toweather reports and phone solicitations generated by computers. Butcomputer-generated speech still sounds quite strange, and one wouldnot mistake it for the human-generated variety for long. Moreover,these systems are incapable of improvising away from their setscripts concerning barometric pressures and the advantages of a newvacuum cleaner.

    This poor record contrasts with scientists? much greater success inprogramming computers to play chess. Another of HAL's accomplishmentsin 2001 was beating the human crew members at chess?aprediction that has turned out to be entirely realistic. We usuallythink of chess as a quintessentially intellectual activity that can bemastered only by the best and brightest. Any ordinary person, in contrast,can talk your ear off in understandable English without necessarilybeing regarded as intelligent for doing so. Yet althoughcomputer programs can now beat the best chess players in the world,no artificial system exists that can match an average five-year-old atspeaking and understanding English. The ability to speak and understanda human language is thus much more complex in objectiveterms than the tasks we usually consider to require great intelligence.We simply tend to take language for granted because it comes soquickly and automatically to us. Just as Navajo proved harder thanother codes during World War II, so English proves harder than theNimzowitsch variation of the French defense in chess.

    The experience of computer science confirms not only that humanlanguages are extremely complex but that they differ in their complexities.Another major goal of artificial intelligence since the 1960shas been machine translation?the creation of systems that will takea text in one language and render the same text in another language.In this domain the ideal is set not by HAL but by Star Trek: All crewmembers have a "universal translator" implanted in their ears thatmiraculously transforms the very first alien sentence it hears into perfectEnglish. Again, real machine translation projects have provenmore difficult. Some programs can take on tasks like converting theEnglish abstracts of engineering articles into Japanese or providing aworking draft of a historical text from German in English or translatinga page on the World Wide Web. But the products of these systemsare very rough and used only in situations where an imperfectaid is desired. Indeed, sometimes they make embarrassingly funnymistakes. Harvey Newquist reports an apocryphal...