Babel No More

by Michael Erard

  This was before I’d learned enough about hyperpolyglots to generalize about them, so his statement seemed rather bold to me at the time. It was more sensible to assume that they popped up randomly in the population, at the same frequency as other types of eccentrics. I thought, he knows more than I do. Maybe he’s right.

  Something else he wrote stopped me cold.

  “On the whole one should concentrate on modern polyglots, usually born after 1900,” he wrote. “That means that Mezzofanti must never be mentioned; he has nothing to do with polyglottery as a science—may be regarded as a mythical person.”

  A mythical person? This didn’t make sense. I’d seen Mezzofanti’s handwriting, his papers, his letters. Did Gunnemark know something about Mezzofanti that I didn’t?

  I would never get an answer. By the time I wrote him back to ask, the Swede had passed away after an extended illness.

  Chapter 7

  The one language accumulator that Dick Hudson knew about, Christopher, isn’t someone you can call up for an interview. His answers tend to be monosyllabic, and he’s not known as much of a conversationalist, whether in English (his mother tongue), or in his twenty or so other languages.* Brain-damaged, Christopher can’t do simple self-care tasks; he could never hold a job. His performance IQ is very low (he’s never scored higher than 76) and has a mental age of nine. Remarkably, though, he has a knack for learning languages.

  He spends much less time at languages than one would think. “He spends most of his waking hours digging in the garden, working in the [wool] carding shop, reading newspapers, listening to music and indulging in a variety of other occupations,” wrote University of College London linguist Neil Smith and collaborator Ianthi-Maria Tsimpli, at the Aristotle University of Thessaloniki, in The Mind of a Savant: Language Learning and Modularity, their first book about Christopher’s language abilities.

  Perhaps it was Christopher’s mother’s bout with German measles early in her pregnancy. Perhaps it was the bad fall she had. Perhaps it was her long labor, during which the nurses sent for oxygen. But in February 1962, six weeks after he was born, Christopher was diagnosed as brain-damaged. At six or seven years old, his fascination with foreign languages was triggered by watching the 1968 Mexico Olympics. He often pretended to be from a foreign country, wearing a towel as a turban or a matador’s cape. “He had a precocious talent, but he was also afflicted with a minor speech defect, poor eyesight, and a degree of clumsiness that seemed to confirm the diagnosis of mental handicap,” wrote Smith and Tsimpli.

  Though he’s never been formally diagnosed with autism, he displays many autistic traits. From the documentary films and photographs I’ve seen, he’s a short, very sweet-looking, shambling man with thick eyebrows and a dark moustache. Smith and Tsimpli describe him as “socially unforthcoming” and “emotionally opaque.” He is uninterested in social formalities and unable to perceive what other people might be thinking.

  Another trait of the autist is an obsessive interest in one topic: in Christopher’s case, it’s languages. In French and Spanish, Smith says that Christopher is fluent, and “quite fluent” in Greek. German is easy for him, as is Dutch. Before his appearance on a Dutch television show, someone suggested that he might improve his “rudimentary” Dutch, so after a few days with a grammar and a dictionary, he was able to converse with people before and after the program, performing as so many linguistic virtuosos before him had. In his other languages, he’s accumulated huge amounts of vocabulary, which he learns with what appears to be a voracious, bottomless memory. One time Smith and Tsimpli gave him an hour-long grammar lesson in Berber and left him some basic books. A month later, he remembered everything about the Berber lesson.

  What’s also amazing is that Christopher can also switch among all of his languages quite deftly and translate to and from English with ease. (Going between other languages is harder for him.) The point is, he’s not simply a memory prodigy. Yet when you read more deeply about what he can and can’t do, it’s not clear that he contributes to an understanding of what language-learning talent might look like.

  It depends in part on how you define talent. If it means that someone learns new languages quickly, and also pays attention to the declensions and inflections of words and making the elements in noun and verb phrases agree, then Christopher has a talent. On the other hand, if by “talent” you mean performing anywhere close to what native speakers do, he doesn’t qualify. In many of his languages, much of what he says is repetitive; his translations are imperfect, particularly in languages that are further from Romance and Germanic. As Smith and Tsimpli acknowledge, though he’s fluent in just a few of his languages, Christopher doesn’t know any of them to a native-like degree.

  Others have noted that his English, though that’s his mother tongue, isn’t native-like, if you consider understanding metaphors to be part of a native’s skill set. Phrases like “standing on the shoulders of giants” confound him. Some scientists have suggested that in order to comprehend a metaphor, you have to know that someone intends a meaning other than a literal reference to giants’ shoulders.

  Christopher also doesn’t learn new grammars completely. After four years of Greek lessons, for instance, he grew his vocabulary, became more fluent, and reduced his errors. Yet he didn’t get better at distinguishing syntactically good from bad sentences. Normal second-language learners have a harder time with the structure of words than with the structure of sentences; Christopher was the opposite. He’s genuinely obsessed—and genuinely excellent at—the mechanisms of constructing words, especially their spellings. He can also use four writing systems. Given a newspaper in one of his languages, he selects words and identifies parts of speech and other properties.

  The biggest limitation—and the one that may provide the true measure of what he can do—is that in most of his foreign languages, his English grammar influences what he says or translates. When asked to translate, “Who can speak German?” he answered with “Wer kann sprechen Deutsch?” not “Wer kann Deutsch sprechen?” It’s as if his English is just ventriloquizing in other languages.

  As Smith and Tsimpli put it, “It is not too inaccurate to suggest that Christopher’s syntax is basically English with a range of alternative veneers.” The technical word is a calque—a word-for-word translation from one language to another. (One well-known example is the English calque “long time no see” of the Mandarin hao jiu bu jian.) In languages like Dutch and German (with a syntax and a word order very close to that of English), the learner can calque away and remain intelligible; the further a language’s word order gets from English, the bigger impediment calquing becomes to being understood (and not sounding absurd). This echoes a folk sentiment about polyglots in Norwegian that someone told me: Hvis en nordman hævder at han snakker syv språk, så er seks af dem norsk. (If a Norwegian claims to be able to speak seven languages, six of them are Norwegian.)

  Recently Smith, Tsimpli, and some colleagues gave Christopher another kind of linguistic workout by teaching him British Sign Language (BSL), an experiment they document in their second book about him, The Signs of a Savant: Language Against the Odds. A sign language is an intriguing challenge. Where a speaker produces a strict string of sounds, a signer often assembles multiple bits of meaning in the same moment of time. This makes calquing of the sort that Christopher does in other languages more obvious. Also, a system requiring manual dexterity and eye gaze was a challenge for a man with limits in each.

  Christopher’s abilities were compared to those to a group of university language students who had tested well on a written test of grammar in foreign languages. His understanding of BSL was good. Yet he had difficulty bringing his talents, whatever they are, to BSL. He had difficulty learning to use grammatical functions that required precise hand control. He developed eye contact with his teacher, but because he doesn’t generally look at faces, he misses the facial movements that BSL uses to signal negation or ask questions like “What?” or “Whe
re?” The normal learners were about as good as Christopher was in BSL.

  From the beginning, Smith and Tsimpli hadn’t been interested in hyperpolyglottism itself, so they didn’t nail down in what way Christopher might be talented (at least in spoken and written languages). The theoretical fish they looked to fry concerned “modularity”—the idea that language is a separate function in the brain. It is thought to be so separate that someone with asymmetric cognitive faculties (like Christopher) could have mostly intact language. Modularity was interesting in itself; so was the nature of thought, and whether or not language could be unique to humans or not. Such claims about modularity were a red flag for critics, who dove at Smith and Tsimpli’s conclusions about Christopher in the earlier book. Debunking the analysis of Christopher would take down the argument about modularity, too.

  Some weren’t convinced that Christopher had, in fact, an intact language faculty, because he was so limited in English. (Smith argues that the structures of English in Christopher’s head were intact, which meant he is still a good example of modularity. His communication failures aren’t governed by what Smith locates in the “language faculty.”) Other critics disputed that Christopher qualified as a savant, or that his fascination with languages was so remarkable. A survey of autistic savants from the 1970s showed that 19 out of 119 children were similarly obsessed with language forms. The survey turned up one child who sounded like Christopher, described by his parents as someone with “a working knowledge of French, Spanish, Japanese, and Russian—knows at least the alphabet and pronunciation of Arabic, Hebrew, and several others.” (Many more people on the autistic spectrum are fascinated by machines, however, than by languages.)

  Others commented that Christopher didn’t have a linguistic ability as much as a powerful talent for recognizing patterns. Only coincidentally was this proficiency attuned to language, specifically words. He was also boosted by a powerful rote memory, considered one hallmark of autistic savants (said to be true because they remember what they see or hear without consciously thinking about what they’re taking in. Overall, Christopher’s memory profile, especially in working memory, is very strange). Both together give him analytic ability as well as recall and performance. Yet, because he makes so many mistakes and can’t break away from English sentence patterns, “it would appear that Christopher is not so much a successful learner of languages [in the strict sense] as he is an obsessive accumulator of minutiae that happen to be linguistic in nature,” wrote one reviewer of The Mind of a Savant.

  In the second book about Christopher’s BSL, Smith and his coauthors appeared to agree with that assessment. They now wrote that his abilities “are only partly linguistic” and that his case “provides no relevant evidence” for language talents. Inexplicably, they continued to describe him as “mastering spoken languages” and as “supremely gifted at learning new languages.” They disputed the notion that Christopher is merely a good pattern recognizer—he fails to recognize them in music and games. And he got stumped when trying to learn an artificial language with a word whose meaning depended on where it was placed in a sentence.

  What does this remarkable man have to do with the rest of us? To Smith and Tsimpli, his case means that language learning doesn’t require some traits we take for granted, such as good general learning abilities, average cognition, and a theory of mind. Critics suggest that the only people who would want to perform like Christopher would be those who’d be satisfied with calquing their mother tongues—even though many language learners would happily accept his memory and skills at parsing and assembling words. This shows why language learning isn’t purely a memory feat: you have to make word orders more automatic than you can consciously retrieve through, say, a mnemonic.

  With such profound asymmetries, Christopher couldn’t be the exemplar of a talented language learner. Certainly not the hyperpolyglot. Yet his case forecast a lot about what I should expect to find at the upper limits of language learning: superior abilities in one area accompanied by deficits in others, mainly. That imbalance might be in intellectual skills. It could also be in areas of language itself—good with words and word forms, Christopher was relatively bad at syntax. Yet it also showed that eroded social and pragmatic skills would be a bigger handicap than poor syntax. A foreign-language speaker who calques all the time can bridge communication gaps in other ways. Even if your pronunciation isn’t perfect, if it’s good enough to clarify and repair, then you’re communicating.

  I could not escape the notion there was someone Mezzofanti-like out there. My search had to continue.

  Erik Gunnemark, who was eighty-nine when he died, had written The Art and Science of Learning Languages, a very nearly complete handbook about how to become a polyglot. In the foreword he even quotes a black magic spell to learn many languages: “Catch a young swallow. Roast her in honey. Eat her up. Then you will understand all languages.” Did Gunnemark roast swallows? Probably not. He seemed like someone who took his own practical advice, which was to build activities for learning around concentration, repetition, and practice. He called these the “three pillars” of language learning.

  Many people have an aptitude for doing this that may not emerge until adulthood, Gunnemark wrote. That’s just as well: he considered them better language students than children, because they’re more disciplined and make better use of time. “A child has to learn about the world and a language, an adult only has to learn about a language,” he wrote. Children have the advantage of living in a world mainly composed of concrete objects that can be named by caretaking adults who are also patient listeners. Adults, who dwell in realms of abstractions, could compensate by observing languages in the environment.

  Intriguing ideas. Sensible, even familiar. However, they provided no clue as to why he had told me that Mezzofanti was a myth. Had he read Russell or Watts? Had he found them overdazzled by thin shreds of evidence? Their judgment stretched too thin?

  Gunnemark’s rejection was overprudent, I thought. As packaged in spectacular display as Mezzofanti was, there was some genuine feat there. The cardinal did accumulate many languages, I would have said to Gunnemark, and a good number to a very high standard, and all to the degree that he needed them in the sphere he worked in, even if he couldn’t, in many of them, match what a native speaker could do.

  But, okay, say I ignore Mezzofanti. Who was the best candidate for the world’s most accomplished hyperpolyglot? Gunnemark’s letter to me contained a pantheon of polyglottery, a list of “modern superpolyglots”:

  Eugen M. Czerniawski (b. 1912)

  Ziad Fazah (b. 1954)

  Arvo Juutilainen (b. 1949)

  Donald Kenrick (b. 1929)

  Emil Krebs (1867–1930)

  Lomb Kató (1909–2003)

  Pent Nurmekund (1905–97)

  Why these seven? Why only one woman, Lomb Kató*? Gunnemark didn’t explain.†

  I’d first encountered Lomb Kató, a Hungarian hyperpolyglot, in a 1996 article by Stephen Krashen, a University of Southern California linguist and an expert on language acquisition. In 1995, Krashen had been teaching in Hungary, where Lomb lived and was locally famous. She’d also written a memoir, Így tanulok nyelveket, or How I Learn Languages, which was first published in 1970—only in Hungarian. When Krashen interviewed the eighty-six-year-old, she was learning her seventeenth language, Hebrew.

  Lomb Kató. (Courtesy of Lomb Janós)

  “What Lomb Kató had was a heroic drive to get comprehensible input and to retain it,” Krashen told me, holding her up as a triumph of dedication. She worked hard; she wasn’t afraid of failure; she read prolifically.* What’s also notable about her is that she took her place in an area dominated by men. Lomb’s case supported Krashen’s theory of second-language acquisition, especially the central part of it, the “comprehensible input hypothesis,” which he first proposed in the mid-1970s. Krashen argued that people can both subconsciously acquire and consciously learn language, but that the former is more impor
tant. Acquisition happens, he said, when we understand what we read or hear—not when we speak or write it, memorize vocabulary, or study grammar. This is how children get their first language—they “acquire” it. So do massive multilinguals like Lomb.

  In 1941, Lomb and her husband moved into a room whose previous Russian tenants had left behind trashy romance novels. Using her beginner’s Russian (pried from two ratty dictionary volumes), she began deciphering them, improving her Russian to the point that she could read literature like Gogol’s Dead Souls, a book she had to sew into a Hungarian encyclopedia in order to disguise it from disapproving eyes. Later, she started Spanish by reading a translation of Gentlemen Prefer Blondes—she didn’t read solely classics, in other words.

  Krashen liked Lomb because she didn’t claim to have a talent for doing what she did. “Her last words to me changed my life,” he once said in an interview. “‘Stephen, you are so young. So many years left, so many languages to acquire!’ (I was fifty-four at the time.)” Afterward he “plunged” into languages—reading novels in French, German, and Spanish. His admiration aside, more would later come to light both about Lomb’s methods and about the extent of her abilities.

  In 2008, when Lomb’s How I Learn Languages became available in English, non-Hungarians could finally read about the famous Hungarian for themselves. Lomb’s prose is spry, often sardonic. “One of my goals in writing the book was to remove the mystical fog surrounding the idea of an ‘innate ability’ for language learning,” she says in an early eye-catching passage. “I want to demystify language learning, and to remove the heroic status associated with learning another language.” What makes one successful, she says, is interest driven by motivation, perseverance, and diligence. One important point seems to be to pump yourself up: “Be firmly convinced that you are a linguistic genius.”