Authors: Michael Erard
Possible explanations for talented language learning fall into two general areas. One view says: What matters is a person’s sense of mission and dedication to language learning. You don’t need to describe high performers as biologically exceptional, because what they do is the product of practice. Anyone can become a foreign-language
expert—even an adult. In fact (the story goes), language learners run the gamut, and the successful ones represent the very, very successful end of this spectrum. Their native languages may be as jealous as anyone else’s, but somehow these people aren’t held back from hearing and producing new sounds, words, and grammatical patterns. Believing that language learning isn’t easy and
takes work, they commit themselves to using their time efficiently.
The other view says: Something neurological is going on. We may not know exactly what the mechanisms are, but we can’t explain exceptional outcomes fully through training or motivation. C.J. came to play an important role in this view, because Loraine had measured in him the cognitive features that support quick, easy foreign-language
learning by adults. Presumably these features are more genetically determined than others; though trainable, they seem to be improvable only within certain margins. In time, C.J. would appear as a case study in other people’s work, including linguist Peter Skehan’s at the Chinese University of Hong Kong. Skehan suggested that what’s so special about C.J. “seems to be the capacity to deal
with large quantities of material to be memorized quickly and easily.”
Both Christopher and C.J. have talents that aren’t centrally about language at all, Skehan argued. Rather, they possess cognitive abilities that happen to be very well suited for learning languages. That is, they can recognize patterns and remember learned material. These skills are suited for languages, which are “relatively
simple codes which can be learned and operated quickly, and which then can be the basis for the retention of material.” When Skehan wondered “whether there can be an exceptional talent for learning languages, qualitatively different from high [linguistic] aptitude,” his answer was a resolute yes.
Both types of language learners would share a couple of traits. They “would have a high range of
lexicalized exemplars, considerable redundancy in their memory systems, and multiple representations of lexical elements. . . . It is assumed that such learners would not value form highly,” he wrote. To translate: they know a lot of words, have many different words for the same meanings, and do not care too much about avoiding errors. People like this have to be biologically different, Skehan suggested.
“Exceptionally successful foreign language learners consistently seem to be characterized by the possession of unusual memories, particularly for the retention of verbal material,” he wrote. “Such exceptional learners do not seem to have unusual abilities with respect to input or central processing.” In other words, they learn languages in virtually the same way everyone else does; they just
have better memory retention and retrieval.
What does better retention mean? The average person, over a lifetime,
will find it easier to learn new facts than to learn new motor or cognitive skills. In the same way, the average adult language learner will have an easier time picking up new words than picking up grammatical rules. And we know, at least in the early stages of foreign language learning,
average adults lean quite heavily on “declarative memory.” This is the part of the memory system that remembers facts and words. It remains fairly robust as one ages, though the plasticity of procedural memory, where motor and cognitive skills as well as grammatical rules are stored, becomes less assured. So “better retention” may mean that hyperpolyglots can cement declarative memories in
their brains more quickly. Or that their procedural memories remain plastic longer than most. Or maybe they just have declarative memories with a lot of capacity. The degree to which Christopher, the polyglot savant, has superior memory retention suggests that he’s less like an average adult language learner than he might at first seem.
For her part, Loraine thought something more might be going
on.
C.J. forgot images and numbers as fast as anyone else did. Why should someone have a good memory for sounds and words but not for other things? In music, he was average; tests of visuospatial ability stumped him; he said he couldn’t read maps or figure out new routes. This intrigued her, since it’s often thought that exceptional verbal abilities are associated with limited visuospatial abilities,
or vice versa.
To give these overlaps some order, she looked to a complex theory known as the Geschwind-Galaburda hypothesis, which links co-occurrences among dyslexia, gender, handedness, and other traits. As examples, there’s a predominance of left-handers among talented visual artists, and males are overwhelmingly more often dyslexic and autistic. In the 1980s, neurologists Norman Geschwind
and Albert Galaburda looked to brain development for an answer. They observed that the left hemispheres of the brains of fetal rats developed more slowly if testosterone spiked at certain developmental moments. The cells destined for the left hemisphere migrated to the right hemisphere, which acquired more of the raw materials for building dense brain connections.
If the same thing happened in
humans, Geschwind and Galaburda suggested, that asymmetry could create clusters of talents and deficits. Their theory might explain why children with left-hemisphere-related
disabilities (such as dyslexia or stuttering) tended to have higher-than-normal abilities in right-hemisphere abilities, such as putting together puzzles. And it may explain why left-handedness (or ambidexterity), homosexuality,
autoimmune disorders (such as asthma or allergies), learning disorders, and talents in music, art, and mathematics all seem to happen together, if not in the same individuals, then in families.
C.J. fit the profile. He was an identical twin, though his brother had no apparent special abilities; neither was strongly right-handed (C.J. was left-handed and his brother was ambidextrous); C.J. had
hives and allergies; he was homosexual; and he got lost easily.
Doesn’t it disprove the theory that C.J.’s brother had no language abilities and wasn’t gay? No, said Loraine. “Just because you have a lot of left-handers with no talents, that doesn’t hurt the explanation, because these cluster in families,” she said.
You’re not looking for a specific gene for deficits or talents; you’re looking
for genes that affect how hormones work. According to Geschwind and Galaburda’s theory, fetal genes drive the production of testosterone and determine vulnerability to hormonal spikes. Neither genes nor hormones determine each other or any particular outcome. Rather, they take a zigzag course: the right inputs at the right moment under the right conditions might produce a certain brain asymmetry
that would lead to behaviors that, culturally speaking, would be interpreted as talents or deficits. Whether the hyperpolyglot learned languages very quickly, could use a lot of them, or both, his or her talent was the result of intersecting lines of circumstances.
As any good expedition begins with a meal, Loraine and I considered hyperpolyglots over sushi. Tomorrow we’d be visiting the brain
institute, and we had much to discuss.
Even if one doesn’t agree that hormonal disarray creates a cluster of traits, the clustering is still of interest. Loraine wanted to know if any of the people I’d met manifest attributes of the Geschwind-Galaburda cluster. I told her that most of the hyperpolyglots seemed to be men, and a number were gay and left-handed. Alexander couldn’t drive. He’d taught
himself to write with his left hand, and his father was an identical twin. I also told her about a translator I’d met at the European Commission in Brussels, forty-three-year-old Graham Cansdale, British by birth,
who came from a “resolutely monolingual family” (as he put it), yet had studied, or dabbled with, a total of twenty-two languages (including Guarani, the indigenous language of Paraguay,
and Vietnamese, which stumped him); he uses fourteen of them professionally (French, Spanish, Italian, Swedish, Russian, Portuguese, Hungarian, Danish, Greek, Czech, Slovak, Arabic, Turkish, Finnish). When I met him he was studying Arabic, Chinese, and Turkish in his spare time.
“People think it must be lots of hard work,” he told me. “In my case I couldn’t agree. I don’t spend a vast amount
of time doing this. It comes so easily. If it is hard work, it doesn’t feel like hard work. Actually, I just remember it because I do. It just goes in. I’m not forcing things in my head.”
Graham was unable to learn how to drive and had an excellent memory for language-related things but not for, say, history facts. And he was gay, married to a Slovakian man. He wasn’t a word accumulator but a
system builder; he was content with the geekish thrill of being able to decipher the verbs on the small print of a menu or to read the street signs that he encountered on his extensive global travels.
This was an anecdotal sample, but the Geschwind-Galaburda hypothesis helps make sense of it. As a theory of both biology and society, one that accounts for individuals as well as families, Geschwind-Galaburda
explains a lot, and has the virtue of not claiming that people with traits A and B are bound to become a Y. It doesn’t indulge a crude determinism but instead tries to map a set of possible influences. Another virtue is that it treats handedness, immune disease, talent, and even sexuality as falling along a spectrum.
Yet its assets—and, indeed, its power—also make it difficult to establish. To
dig into it, you’d have to know genetics, cognitive science, epidemiology, and endocrinology, all in great detail. You’d also have to gather a big sample, many families, and ask them the right questions. Attempts to confirm some of the correlations have had mixed success. One study, for instance, found no connection between left-handedness and higher spatial or mathematical abilities, though researchers
did find that left-handers often reported having speech problems. Various studies have shown that people with autism have higher rates of non-right-handedness. Yet another study found, as the Geschwind-Galaburda
hypothesis would predict, that there are more males than females who perform both very high and very low on mental rotation tasks, though, outside of Loraine’s work with C.J., no one had
ever linked verbal abilities in a foreign language with any of the Geschwind-Galaburda traits.
We discussed how I might capture similar information about clusters of traits from as many people as possible, and I eventually designed an online survey that collected information from nearly four hundred people from around the world from January 2009 to January 2010. People who claimed to know six
or more languages were directed to the survey via English-language linguistics blogs and language learning websites like
http://how-to-learn-any-language.com
. There they gave their consent to answering questions about their background, their language learning, and their cognitive styles.
*
The results, which were analyzed by an academic statistician familiar with this type of data, give support
to some parts of the Geschwind-Galaburda hypothesis. For instance, people who reported knowing six or more languages and who said that learning foreign languages was easier for them were more likely to report homosexual behaviors, preferences, and/or orientations than would be predicted. This finding was statistically significant.
This same group also was more likely either to have immune diseases
themselves or to have family members who did. However, neither of the other traits (handedness and twinning) had any meaningful relationship with either number of languages or ease of learning.
†
Despite an anecdotal connection between homosexuality and verbal abilities, no connection between language talent and homosexuality had ever been shown in research before. Again, this isn’t saying that
people who speak a lot of languages or learn them easily are necessarily gay; it’s that there are more gays—and people with immune diseases, for that matter—among
talented language learners than you’d otherwise predict. Testing this with numbers allows us to move beyond anecdotes and to suggest why these patterns exist.
This survey’s groundbreaking results were many months away; Loraine and I
were still sitting in the sushi restaurant, preparing for our visit with Krebs’s brain. She said, “There’s one difference between C.J. and people like Mezzofanti, Krebs, and the other people you’re interested in. C.J. hadn’t started to accumulate languages when we met him, and we don’t know how well he switched among them. Rather, he was notable for how fast and easily he learned his languages to
a very high level. I’m interested specifically in those people,” she said. They may or may not overlap with those who
accumulate
languages. The question was, are they the same? Or are there two separate groups?
I repeated a phrase that had been coined by someone named Loren Coleman, who writes and comments prolifically about the search for hidden or unknown animals, from Bigfoot to newly discovered
frog species. Coleman once summed up such uncategorized creatures, also known as cryptids, as “out of time, out of place, out of scale.” In the case of the coelacanth, they’re living fossils, thought to be extinct, so they’re out of time; they’re out of place, like a monkey escaped from the zoo who is sighted rooting in suburban trash cans; or they’re out of scale—like giant Amazonian snakes
or mastodons. The term fits hyperpolyglots perfectly—they do things with language that are out of scale with what normal people do with language. They don’t usually live with other hyperpolyglots, so they’re also out of place. Some of them bear the distinct whiff of the past or are perfumed with the future. That takes them out of time. C.J. and Mezzofanti, Alexander and Helen, Ken Hale and Christopher—all
of them out of time, out of place, out of scale. I felt lucky to have found them in their place, and to have seen the scale you’d need to measure them.