Authors: Michael Erard
what World offers is vain
Oh let us Heaven gain!
There was also this poem, which Mezzofanti wrote in Italian:
Of all the thousands of voices in various accents
that come from human breasts in hundreds of languages
the one that’s dearest to a virtuous and modest heart
is the voice that praises and extols the Creator.
Perhaps because he’d staked his sense of self on languages, his ego, vulnerable to accolades and attention, needed dampening.
As I was leaving the library’s main doors, Pasti turned to me to say that he doubted Mezzofanti’s skills. This surprised me. In his book, he hadn’t probed
Mezzofanti’s reputation with any skepticism. Now he was telling me that back then, the standards for what it meant to speak or know a language were “very low.” “I don’t think he actually spoke Persian or Arabic.”
“How do you know?” I asked, hoping he’d uncovered some damning evidence.
He shrugged. “Just a feeling,” he said.
Knowing I might never visit the archives again, I had hit on a solution
to get at Mezzofanti’s proficiency: I’d count the letters he received in each language. If he got many, he must have been writing a lot, and that, maybe, pointed to a great deal of practice, then to a high degree of proficiency. It was a fair social science hypothesis.
I told Pasti about my plan. The librarian smirked at me.
“You are a positivist, I think,” he said. A positivist is someone who
believes you can get at truths only through what can be counted, measured, and observed. I was shocked—I’ve been called names before, but never
that
. Pasti had interpreted Mezzofanti’s life, leaving tabulation to people like me. With a grin, the librarian punched my arm lightly.
“I have to leave Bologna with some hard facts about Mezzofanti,” I said. “People are going to expect me to come up
with a hard answer. I just can’t give them more anecdotes.”
Pasti’s smile was unforgiving. “That proves it,” he said. “You are a positivist.”
Back in the Archiginnasio, I slumped at the table with an unopened box and the
inventario
in front of me. Pasti wasn’t the first person to
discourage me about Mezzofanti’s true nature. And he wouldn’t be the last. Maybe he didn’t quite understand my desperate
desire for a visual encounter with the hyperpolyglot, if not in flesh, then in fact.
One thing was clear: I could stay in Bologna until my Italian was
molto perfetto,
and the truth about the hyperpolyglot would elude me.
I departed Bologna with one intact conclusion: Mezzofanti learned and used numerous languages in a range of ways.
Given the rich purpose evident in those boxes, I was entirely
willing to extend to his legacy the notion that those uses were real and meaningful. Even if many of them were narrow, they were
his,
no less legitimate for being idiosyncratic. I also felt certain that the languages weren’t objects of beauty that he displayed like butterflies in a collection. Though I couldn’t read most of the languages, the artifacts indicated that he
used
them. Maybe not all
at the same time—some he might have kept on the back burner. To my eye, at least, they looked like tools. Some of what he used languages for clearly involved high-stakes communications. Some of it was the equivalent of a verbal
amuse
-
bouche
. Some of it was memorized chitchat (“Will you visit the Pope in Rome?”) for social lubrication. His abilities may have been less than perfect or fluent or
native-like. That patchwork quality was something I would find to be common among hyperpolyglots.
My skepticism eroded on another front, as well. Without some kind of mental gift, Mezzofanti couldn’t have made so much—even half as much as was claimed—of the language encounters that he sought. After all, many people live at linguistic crossroads where they can be immersed in many languages; some
may even love languages. Yet very few become hyperpolyglots. Moving forward, one crucial dimension of my search had to be to characterize, as best I could, the mental powers that he and other hyperpolyglots brought to bear on language learning.
I also recognized that his linguistic patchwork reflected a level of brainwork that wouldn’t be captured in a single number representing the languages
he spoke, read, or translated. One skepticism I was willing to jettison: disqualifying someone solely because they had a large number of languages associated with their name. Other things could
disqualify them, I vowed. Not the number of languages. (Nor could it
qualify
them, but more about that later.) I also vowed that I wouldn’t look at people with an “all or nothing” eye but only a “something
and something” one. Among other things, this meant that individuals could be interesting even if they weren’t native-like in all their languages.
I left with the belief that Mezzofanti did things with languages that the people who speak them natively would never do. Topping the list was his ability to rapidly analyze languages and his prodigious memory (as evidenced by performance and his own
claims); an apparent ability to mimic speech sounds that weren’t native to him; and an ability to switch among his languages without interference. These are unique skills; monolingual native speakers don’t necessarily have them, and, except for the language switching, neither do bilinguals. The native speaker wasn’t the hyperpolyglot’s twin, joined in comparisons; someone else would have to take
that role.
Mezzofanti’s time and place seemed distant and inaccessible. But was he truly one of a kind, the only member of his species? Or could others like him be living among us now?
Part 2
APPROACH:
Tracking Down Hyperpolyglots
Chapter 5
N
earing retirement, Dick Hudson, a linguist at University College, London, took up an overlooked query: Who had learned the most languages ever? In the mid-1990s, he sent this question to LINGUIST List, a popular forum with language scientists. A flurry of postings listed the names of well-known hyperpolyglots of yesteryear, including Giuseppe Mezzofanti. Others cast doubts that the
upper limit for languages would be very high. For all purposes, it remained an open question, a natural experiment whose results had never been analyzed.
A couple of years later, he received an email that began like this: “Sir, First, let me apologize for bothering you, but I saw an article you wrote and had to write.” The writer, N., had found Hudson’s posting and wanted to describe how his
grandfather, a Sicilian with no formal schooling, had learned languages with such remarkable ease that by the end of his life he could speak seventy, as well as read and write in fifty-six.
*
This grandfather was twenty when he moved to New York in the 1910s. There he found a job as a railroad porter, which brought him into contact with travelers who spoke many different languages. N. said
he once
watched his grandfather translate a newspaper into three languages on the spot.
When N. was ten years old, in the 1950s, he accompanied his grandfather on a six-month world cruise. Whatever port they called at, N. said his grandfather knew the local language. Their trip took them to Venezuela, Argentina, Norway, the United Kingdom, Portugal, Italy, Greece, Turkey, Syria, Egypt, Libya, Morocco,
South Africa, Pakistan, India, Thailand, Malaysia, Indonesia, Australia, the Philippines, Hong Kong, and Japan. Assuming the grandfather spoke the local language at each port, one can figure that he knew some of at least seventeen (including English)—though there was no mention of what he could accomplish in each.
Even more amazingly, N. claimed that this talent ran in his family. “Every three
or four generations there is a member of my family who has the ability to learn many languages,” he wrote. His grandfather once told N. that his own father and great-uncle could speak more than one hundred languages.
When Hudson read this, its significance was crystal clear. Studying the genetic basis of abilities in language, especially the heritability of language disorders, is cutting-edge
territory. In the 1990s, exciting work was performed on a family with developmental language problems who had the same mutated gene. Could there be a genetic link for exceptional abilities? Perhaps a hyperpolyglot gene?
With N.’s permission, Hudson passed the mail on to LINGUIST. Seeing a need for a better label for people who learn lots of languages, Hudson chose
hyperpolyglot
.
Multilingual
didn’t cut it;
polyglot
felt too pedestrian. A hyperpolyglot, in Hudson’s terms, was a person who could speak six or more languages. Hudson had found that community-based multilingualism, where everyone, not just special individuals, spoke many languages, had a ceiling of five languages. So someone who spoke six or more had to be exceptional, Hudson reasoned. They weren’t just polyglots, they
were
hyper
polyglots.
Hyperpolyglots caught Hudson’s attention because they were beyond the pale of linguistic theory. With only one exception that he knew
of—the study of a severely mentally impaired man named Christopher who knew twenty languages—no one had ever addressed the issue. No longer a professor affiliated with an institution, Hudson had nothing to lose by following up a fascinating
but ignored phenomenon, though N. disappeared shortly after his letter was published.
If you believe that language is innate and uniquely human, then the question about how many languages a person can learn would seem trivial. Learning one language is itself an evolutionary marvel; learning many of them, well, that’s gilding the lily. A language superlearner deserves about the same attention
as an ultramarathon runner—accolades and respect for pushing basic human equipment to an extreme, but not the scientific interest you’d give to, say, a human who could breathe underwater or flap his arms and fly.
If pushed to explain the hyperpolyglot phenomenon, an innatist might say that every baby’s brain comes equipped with a primordial universal grammar, a sort of periodic table that contains
the basic properties of all the world’s languages and all the dimensions along which they differ. Coming into contact with a real language (or several) triggers some properties from that table. The ones left untriggered disappear. Because all the languages on the planet differ along a finite number of dimensions, even an adult should be able to retrigger these properties and their permutations,
given enough time. Theoretically, there should be no limits on how many languages a person could learn. If most people don’t, it’s because they’d rather carve whalebone, grow the perfect orchid, or sit on the beach.
Those of an opposing theoretical bent, the emergentists, see language not as inborn but as a behavior that grows out of a set of simpler, overlapping cognitive skills. At certain
moments in a child’s development, such skills are enhanced by experiences that bootstrap them to a higher level of complexity. In this account, protogrammars don’t have to exist in babies’ brains—as expert learners, humans can fill in gaps from their environment. Accordingly, an emergentist’s explanation of the hyperpolyglot goes something like this: people with a gift for learning languages must
recognize and parse patterns extremely well. Surely this makes a kind of intuitive sense. But here the account stalls. At what point does parsing patterns become knowing a language? And how can
you escape the influence of your earliest language—or harness your knowledge of it—to become fluent in later ones?
Meanwhile, the applied side of linguistics, which deals with foreign-language education
and literacy, developed the notion of language-learning “aptitude.” It was really a measure of how quickly a student could achieve proficiency in an extra language in an allotted time period. As the concept was first developed in the 1950s, aptitude had four dimensions—how well a person can recall sounds; how sensitive she is to grammatical patterns; how well she can produce new sentences based
on what she analyzes; and how well she learns how words in the first and second languages connect. Later, the notion evolved, adding new understandings of how memory works.
But aptitude was never applied to hyperpolyglottery. The bulk of research on aptitude had been done by the bureaucrats of the foreign-language education establishment in Washington, D.C. (and other capitals), where the typical
learner of interest was an adult who’d mastered one language well enough for spycraft and diplomacy. The massively multilingual person? Barely interesting.
Even if hyperpolyglots had been examined through the lens of aptitude, the concept had shortcomings. For instance, standardized tests of aptitude tend to predict a student’s grades in certain kinds of classrooms better than they predict real-world
language use. They also don’t predict how adept a person could ultimately become in a language—high aptitude is no guarantee that abilities won’t peak quickly. More important, there’s no agreement on what aptitude is made of, how much it can be cultivated, and how much is inborn. Aptitude could make for a potent political time bomb in many educational settings. Holding out high aptitude implies
low aptitude, which, in turn, implies the promise of handing out educational resources on an unequal basis. If the aptitudes you find striking happen to correspond with certain genders, races, or classes, then you’d look as if you were cementing old privileges. It’s safer to assume that all students possess a generic cognitive profile and teach them with fitting uniformity.