In 1913, at the age of forty-five, Krebs married Amande, a German divorcee. On a honeymoon tour, at a stop at the tomb of Confucius, he read the inscriptions in Mandarin, Manchurian, Mongolian, Kalmuck, and Turkish. Frail and perpetually underpaid, Krebs (or “Krebsy,” as his wife called him) sat down the following year and wrote a list of what languages he could use—he could, for instance, translate into and out of German in thirty-two languages.* Later he would be said to “know” sixty or sixty-five languages. His stepdaughter appended her own note to the list: “It is a great difference between whether one can speak, write, and master a language, or whether one is able to finish correct translations as a proven interpreter.” Be that as it may, during his lifetime he passed government tests in Chinese, Turkish, Japanese, and Finnish, and possibly more.
Similar to other hyperpolyglots I had met or read about, one of Krebs’s most stunning traits lay in how quickly he could learn. Werner Otto von Hentig, a young German attaché in China, described how Krebs had jumped up in the middle of breakfast to find out from two strangers what language, “foreign to him, had been battering his ear.” Armenian, he found. After ordering books, he spent two weeks on the grammar, three on old Armenian, and four on the spoken language. “Then he was a master of them too,” Hentig wrote.
If the anecdotes about Krebs’s language genius—and his obsessive passion—are as rich as those of Mezzofanti, they differ in one respect: they tell of a character who was rude and impatient. Hentig related that Krebs once refused to speak to his wife for three months because she had told him to wear an overcoat in December; in one year, he fired a succession of eighteen Chinese cooks, none of whom pleased him. Once, in order to satisfy a bureaucratic requirement, Krebs had to take a test in both Finnish and Japanese. He intimidated the examiner with his knowledge, scaring the man from the room. In China, Krebs made it perfectly clear that he wanted to study languages rather than do his job (especially since he was often sleeping during the day, having stayed up all night studying).
In another revealing anecdote, Hentig described having to fetch Krebs for a meeting.
“His Excellence wants to see you!” Hentig shouted over the walls of Krebs’s compound. There was no answer. “Herr Krebs, the legate needs you!” No answer. “The Herr Minister is asking for you!” Finally Hentig heard a grumble.
“The legate knows me, leave me in peace,” Krebs grumbled.
“May I help you get dressed?”
“Go to hell!”
“They really need you.”
“They always say that,” Krebs muttered.
One contemporary said that Krebs had never learned the “technique of life.” He was someone who could tell you off in dozens of languages. He had translated the phrase “kiss my ass” (known as the Swabian salute) into forty languages. Something of a joker, he gave a German journalist the Chinese name Bu Zhidao, “doesn’t know.” In daily life he was so disagreeable that no one wanted to work with him, which became a liability later in his career, when no one was willing to promote him or accept his work in languages other than Chinese.
Like Alexander Arguelles, Krebs reviewed his languages on rotation: a strict schedule that assigned Turkish to Monday, Chinese to Tuesday, Greek to Wednesday, and so on. With a book in hand, he walked around and around the dining room table from midnight to four in the morning, naked, smoking a cigar, drunk on German beer. His library was organized by language and language group. For each book he wrote a summary, which he regularly reviewed. At his desk, he stood. He refused to eat anything but meat, and sought out social interaction only if he could use one of his languages. “He knew 32 languages, not in the way we often see with polyglots, but elegantly and well spoken in Arabic as well as Russian or Italian,” Hentig wrote. His Tuscan dialect was so good, the Italian ambassador in Beijing offered to cut Krebs’s hair, just to be able to hear Tuscan.
Unavoidably, writers of the time compared him to Mezzofanti, but in the eyes of German writers, Krebs was superior. “Against whatever kind of displeasing experiences are told of the Mezzofantis, who know all languages, but none fundamentally,” wrote Ferdinand Lessing, a German translator in Tsingtao and later a professor and polyglot himself, “this wonderful talent bites its thumb.”
If they could have roamed freely once they reached New York, the Krebs family would have found a city filled with immigrants from Europe—in fact, there were more daily newspapers published in other languages (Italian, German, Hungarian, French, Croatian, Spanish, etc.) than in English at the time. In addition to twenty-two English newspapers, there were ten Italian, seven German, seven Yiddish or Hebrew, three Greek, three Hungarian, two French, two Bohemian, two Croatian, and one apiece in Spanish, Serbian, and Syrian. Perhaps the Krebs family was detained at Ellis Island, the main through-point for immigrants arriving on the East Coast. There he could have met professional language learners who, like himself, had been employed to deal with the polyglot huddled masses. One of these was an Italian immigrant named Anthony Frabasilis, a renowned scholar of Greek philology at the University of Athens, who was hired at Ellis Island as a Greek interpreter in 1909 and also worked in Italian, Spanish, French, German, Polish, Russian, Turkish, and Armenian. He’s said to have known fifteen languages, and to have spoken them all well.
Starting in 1909, a civil service exam tested interpreters’ writing, reading, and speaking abilities in some (or all) of the languages in which they worked, which provides hard evidence for a cluster of real hyperpolyglots at Ellis Island. Another interpreter employed there was Reuben Volovick, a native of Russia, who knew Yiddish, Russian, Ruthenian, and many other Slavic languages. Another was Peter Mikolainis, a Lithuanian native, who knew seven languages. They would meet with passengers, sorting and directing them through medical exams and legal interrogations. Except in rare medical cases, the interviews were simple enough: What’s your occupation? Your race? Your ethnicity? Have you ever been an anarchist? A polygamist?
Finally, the Krebs family boarded a ship for Europe, leaving behind Emil’s extensive library, which eventually was sold to the US Library of Congress. Back in Germany, Krebs turned to languages with full force, “surrendering to his great ambition for language study,” as his great nephew, Eckhard Hoffmann, wrote. The Foreign Office was offering 90 deutschmarks for every language that someone could speak. “You’ll be a millionaire!” family friends told him. But officials informed Krebs that he would be restricted to testing in two languages. He made nothing for being able to read the cuneiform writings of Assyrian, Babylonian, and Sumerian.
One afternoon in March 1930, while he was translating something (what isn’t known), Krebs collapsed, and he died soon afterward. The news spread quickly, and later that day, his wife received a chilling call: Would the family donate his brain to science? The request came from Oskar Vogt (1870–1959), a pugnacious specialist in brain anatomy and the director of the Kaiser-Wilhelm Institute for Brain Research. The brain would be a fine addition to Vogt’s collection of elite brains, and the only brain of a Sprachgenie.
Vogt, devoted to the study of elite brains, had a trail of scrapes and narrow escapes behind him. In 1924, he’d been invited to Moscow to study the brain of Vladimir Lenin. At the time, political maneuvering in the young Soviet Union, particularly by Joseph Stalin, saw the importance of creating and sustaining a cult of Lenin as a revolutionary supergenius. “It also appeared a brilliant idea to obtain, if possible, a confirmation of Lenin’s ‘genius’ from some respectable source, preferably from abroad,” wrote Igor Klatzo, a biographer of Vogt and his wife, Cécile. By 1927 Vogt had soaked Lenin’s brain in formalin, embedded it in paraffin, and sliced it into 31,000 sections. From there, he faced a huge challenge: how to remain scientifically principled yet not offend his Soviet sponsors? How to explain certain features of Lenin’s brain without referencing the fact that he might have had syphilis? (He didn’t—Lenin apparently had a family history of atherosclerosis.) And what if Lenin’s brain didn’t match other elit
e brains?
Vogt solved the problem by describing the rich array of pyramidal neurons in Lenin’s cortex and explaining that they must have been involved in rich imaginative, rational thinking. To great Soviet acclaim, his paper was published in 1929, and Vogt turned to other projects.
Little did he know he would be collecting a hyperpolyglot brain the next year. He met Krebs’s sister-in-law and his stepdaughter in the church where the funeral was to be held; by law, brain extraction required family members to be present. Toni and Charlotte-Luise, who had stepped away because they couldn’t bear to watch, could hear Vogt’s hammering and sawing. The mood must have been one of Frankensteinian gloom: the dark church; the flickering gaslight; and Vogt walking away with Krebs’s brain, jiggling in a glass jar.
It was this brain, I hoped, that would have something to say for itself.
Chapter 11
The brains that have added the most to our understanding of language abilities belong not to the Mezzofantis or Krebses of the world but to people who have lost those abilities. One of the more famous of these was a man named Leborgne, an epileptic laborer in Paris who had barely spoken for years when he was brought to the hospital in 1861 for an inflammation of the legs. He could only say “tan” (though he could say it with different intonations) and gesture with his left hand. The hospital staff, and later the medical literature, knew him as Tan Tan.
When, several days later, Leborgne died, his doctor, Paul Broca, performed an autopsy and found, on the left front side of Leborgne’s brain, an area of tissue damaged by syphilis. Later, Broca collected the brains of others who had the same broken speech as Leborgne (we now call it aphasia) and damage to the same part of the brain. This led Broca to propose, against prevailing scientific opinion, that he had found the brain’s locus for speech functions. Also contrary to ideas of the day was his proposal that speech control was located only on one side of the brain (in most people, on the left). More confirmation was needed, but the outlines of Broca’s notion soon passed into fact. Now scientists recognize that this area is responsible for more functions than previously thought—and also that languages are controlled in other places besides “Broca’s area.” But it was an important discovery nonetheless.
To get a sense of where exactly Broca’s area is, imagine a globe, perhaps one of the glass or plastic globes you’d find in a classroom or library, mounted on a metal bracket, the sherbety chunks of nations spread across it. My globe happens to be a vinyl inflatable one, which I’m going to use as a rough model for the geography of the brain. To start, turn the globe so that Europe faces you, and the imaginary north–south line, the prime meridian, which runs through Greenwich, England, lines up with your nose. This puts the prefrontal cortex, the seat of human consciousness and cognitive control, in an area that ranges from the equator northward to the north pole, east to the Middle East, and west across the Atlantic Ocean in line with the northern coast of Brazil.
This means that when I place my hands at the equator on opposite sides of the globe and hold it front of me (the prime meridian in line with my nose)—my left hand on the brain’s right hemisphere, my right hand on the left—the edge of my right thumb will be right next to the Arabian Peninsula and the Arabian Sea. This is the area of the brain that Broca associated with speech control. (There’s an analogous area on the right hemisphere right off the coast of Brazil, though for most people it plays a less important role in language. Not for Emil Krebs, as we’ll see.) About thirteen years after Paul Broca made his observations, a German scientist named Carl Wernicke announced that people with damage to another area of the left hemisphere also had communication problems. If you return to the globe, this area is in far eastern Mongolia, or where the end of my right index finger lands.
Broca’s area usually gets the attention when people talk about language. But recent scientific work has revealed a broader story, thanks mainly to imaging technologies like functional magnetic resonance imaging (fMRI), positron emission tomography (PET), and diffusion tensor imaging (DTI). In that broader story about language in the brain, language doesn’t reside in any single place. Instead, it is spread over parts of the brain in networks that are more widely distributed than had been thought. There are two main networks, which have also been called “streams.” One stream maps speech sounds onto strings of motor commands, so that a person can turn sounds they hear into sounds they say. On the brain-as-globe, this stream goes north, from north India to eastern Mongolia. In most people, this stream is located on the left hemisphere of the brain. The other stream maps strings of speech sounds onto stored concepts in the brain, which is how you’re able to comprehend what someone is saying to you. This network seems to spread in both hemispheres, on the left hemisphere from north India to northern Taiwan, on the right from south Texas to Hawaii. Greg Hickok (at the University of California at Irvine) and David Poeppel (at New York University), the two neuroscientists who put together this “dual stream” model in 2000, call the first one the “how” stream (since it tells the system how to produce meaning in speech) and the second the “what” stream (since it metabolizes a stream of speech sound and translates it into meanings).
The “how” and “what” model has been backed up by neuroanatomical studies that have traced the bundles of nerves that connect these areas to each other. As a complete picture of everything linguistic in the brain, it has shortcomings—it’s mainly a model of words, even though human language obviously involves grammar. But it illustrates that languages aren’t filed away like library books in the brain; there’s no “French” neural pathway living next to, or occasionally overlapping, a neural pathway that’s dedicated to “English.” Instead, what you have are meanings with different sets of strings of sounds attached to them, some of which the outside world identifies as French or English. It also helps to illustrate some of what’s going on when languages are learned. Learning how to speak a new language will involve the “how” stream, engaged mainly on the left side of the brain (under my right hand); learning how to comprehend a new language will involve the “what” stream, which is likely working on both sides of the brain (under my left and right hands). Getting those two streams up and running makes up a lot of the work of foreign-language learning.
If this seems very distant from everyday life, I find it extraordinarily useful for understanding why my seventeen-month-old son can understand me better than he can speak, and why he can understand words that he can’t say. It’s because the “what” stream doesn’t involve anything mechanical, whereas the “how” stream uses lots of moving parts. And knowing that his brain is plastic also helps explain why his pronunciation of the same word varies so much from day to day—the motor commands haven’t become concrete yet.
The fact that early languages, no matter how many there are, utilize the same streams implies that the brain doesn’t have a native language. The brain can only reflect the fact that a set of neural circuits was built and activated for a certain period of time. Nor does the brain care if those neural circuits map onto things that the rest of the world calls languages or dialects. It really cares only about what activates those circuits. Thus, the brain patterns that typify language use across skill levels can be mapped.
Brain imaging technology monitors the intensity of oxygen use around the brain—higher oxygen use represents higher energy use by cells burning glucose. The deeply engrained language circuits will create dim MRI images, because they are working efficiently, requiring less glucose overall. More recently acquired languages, as well as those used less frequently, would make neural circuits shine more brightly, because they require more brain cells, thus more glucose.
Using imaging technology to study these circuits, you’d also see the location of oxygen use change, depending on skill level. More recently learned languages engage areas all the way on the other side of my plastic globe, somewhere under my left hand. This is the signature of a brain that’s recruiting higher-level cognitive processes, not automati
c ones, to perform language tasks in relatively new languages. Given enough time and practice, however, those tasks would consolidate into the streams under my right hand. By making expected, predictable tasks automatic, expert brains save up cognitive resources to deal with unexpected, novel tasks.
Building the “how” and “what” streams in adult brains doesn’t happen without conscious practice. But, as Dick Hudson noted, people who practice the same amount vary in their ultimate performance. It makes sense that there must be some sort of underlying capacity that enables those streams to be faster, stronger, or more durable. Because the functional anatomy of the brain is still largely a mystery, mapping that kind of capacity is a challenge—or, to put it another way, just because no structural or anatomical difference has yet been pinpointed doesn’t mean that it can’t exist.
“Neuroscientists will tell you that brains are as different as faces, perhaps as different as bodies,” John Schumann, an applied linguist at UCLA and an expert in the neurobiology of language learning, told me. Partly this is because of gene shuffling at fertilization—you get 50 percent of your parents’ genes but not the same 50 percent as your siblings. The other reason is that the genes don’t determine the exact placement of neurons in the brain. “During embryonic formation, as neurons are formed and migrate to what will be the brain,” Schumann said, “their trajectories are stochastic, and they depend on the chemical and mechanical milieu of the brain.”
One result is that brains are similar in gross anatomy, while at a microscopic level, they are markedly different. Some of those “microramifications” could point someone toward high performance.
Schumann speculated that hyperpolyglots might have different brains than normal language learners. “During embryonic development,” he said, “there’s some neural migration, perhaps to the area between Wernicke’s area and Broca’s area.” This would correspond on the globe to the swath of Central Asia between Saudi Arabia and Mongolia. As a result, Schumann said, there would be more robust formation of brain matter, the neurons and neuropils, dendrites and neuroglia, and support cells. All these I would be learning more about later.
Babel No More Page 17