Becoming Fluent

by Richard M Roberts

  Answers to the tip-of-the-tongue questions:

  1.Sampan (some people think of junk for this definition as well).

  2.Apse.

  3.Nepotism.

  4.Caduceus.

  5.Extortion.

  6.Convene.

  7.Philatelist.

  8.Fawning.

  9.Sextant.

  10.Cloaca.

  Practice Makes Perfect?

  A major part of foreign language mastery is practicing reading, writing, speaking, and listening in the new language. For many, the prospect of intensive practice will evoke memories of rote drills in a high school Spanish or French course. Most students find such exercises to be pure drudgery, and they can drain away all of one’s interest in the subject.

  Practice, however, really is an essential part of this enterprise, even if it seems daunting. It may be helpful to reflect on your goals for learning a foreign language. Remember, to do many things in a foreign language you do not need native-like fluency. For example, if your primary motivation for learning Swedish is to exchange pleasantries with you wife’s extended family, then spending months in intensive study would be overkill. But to even do this, practice is an important part of the story. Our goal in this section is to review the cognitive science research on practice and expertise, and to offer some concrete suggestions about how to make the time spent on practice as efficient and productive as possible. You will learn how to work smarter, not harder.

  What does it mean to be an expert? You may be surprised to learn that the study of expertise is an entire subfield within cognitive science. Researchers have studied individuals from diverse fields of accomplishment in order to understand how expertise is acquired. In the course of writing this book, we’ve spoken to many people who have asserted that they simply aren’t good at foreign language learning; however, they based this perception on negative experiences in high school or college. This kind of reasoning illustrates a commonly held belief—that doing something badly once means to be bad at it forever, creating the conditions for a negative self-fulfilling prophecy.

  One domain of extensive study has been chess playing. It’s easy to identify chess experts, because good players will have a numeric ranking that objectively measures how good they are. There are a variety of such systems (the Elo and USCF systems are two of these), but the basics are the same. If you have a higher rating than I do, and I play and beat you, then my rating goes up a little bit, and yours goes down. Over years of play in many matches and tournaments, a small subset of chess players will achieve titles like “grandmaster.” Many of these individuals have been studied, and it turns out that they’re nothing like what you might expect.

  The common stereotype of the chess grandmaster is someone who possesses very high intelligence and an incredible memory. They are thought to play chess well because they can see dozens of moves ahead into a game. (They might also be thought of as obsessive and antisocial, but those parts of the stereotype aren’t relevant for our purposes.) When researchers began to study chess grandmasters in the 1940s, they discovered that these people often have average intelligence, and normal memory for everything but chess. They also don’t have some master plan for a game consisting of dozens of moves. Because they can’t anticipate every possible countermove by an opponent, it isn’t feasible for them to implement such long-term strategies.

  However, when these experts’ memory skills are tested within the domain of chess, their extraordinary abilities quickly become evident. When briefly presented with a position that had occurred in a real game, chess experts can recreate that position on a second board quickly and accurately, even if that involves the placement of one or two dozen pieces on their respective squares. Novice chess players, as you might expect, perform poorly at this task.4

  How are the chess experts able to perform such feats? An important part of the story seems to be chunking, or the grouping of disparate objects into meaningful patterns, as we discussed in the previous chapter. For example, a very common pattern that occurs over and over in chess games is for the king to be placed behind a row of three pawns. The novice chess player would have to remember this by remembering four separate items in working memory. The expert, by contrast, can remember that pattern as a single chunk, which greatly increases his or her working memory capacity. Perhaps most importantly, both chess experts and novices perform poorly when the board to be recreated has been populated by chess pieces placed randomly. In this situation, the chess experts can’t rely on chunking to aid them.

  Researchers have also tried to estimate how many groupings of pieces a chess expert can recognize instantly. It’s thought that the “vocabulary” of the expert is between 50,000 and 100,000 patterns. It’s also been estimated that it requires about 10,000 hours of study and practice to acquire this vocabulary. Interestingly, this number is consistent with the amount of time required to achieve expertise in a number of disciplines, a finding that has been popularized by Malcolm Gladwell in his book Outliers. However, a recent overview of this literature by Brooke Macnamara and colleagues suggests the effects of deliberate practice vary widely by domain.5

  A belief that deliberate practice is the key has caused some to claim that there really isn’t a fundamental difference between experts and nonexperts. Experts are simply individuals who, because of their intense interest and love of a subject, have put in the time necessary to achieve a very high level of performance. Although this conclusion is controversial, it does provide some hope for those of us who wish to acquire expertise. It’s heartening to think that expertise is ultimately achievable by virtually anyone.

  What’s that? You’re saying that you don’t have 10,000 hours to spare to learn a foreign language? Remember that you don’t need to become an expert to achieve a very useful amount of fluency. But you will still need to practice, and research has shed some light on this topic as well.

  Before reading any further, we’d like you to participate in a simple experiment. We’d like you to estimate how many classmates’ names you can remember from your high school graduating class (first and last names). Perhaps you’re thinking that you can name two-thirds of your classmates, or at least half of them. Now take out a sheet of paper, and try to write the names down. Go ahead, we’ll wait.

  If you actually attempted this experiment, you’re probably disappointed at how few names you were able to produce. You were undoubtedly able to immediately scribble down the names of close friends and well-known class personalities, and perhaps a few others. But most people will find that they’ve created a fairly short list.

  This experiment was first conducted by memory researcher Harry Bahrick. He was interested in studying very long-term memory—the things that we remember across decades of our lives. The difficulty with exploring this topic is that most of us don’t have extensive records to determine whether our memories are complete or accurate. If we were to ask you what you had for lunch on November 5, 2009, you probably have no way of checking whether you consumed a salad or a sandwich.

  Bahrick got around this problem when he realized that almost everyone has detailed records from their late teenage years: their high school yearbooks. He asked a number of individuals to bring their yearbooks to his lab (without looking at them first), and then asked them to carry out tasks like the one you just did: to recall as many of their classmates’ names as they could.

  If you really did attempt our exercise, then perhaps Bahrick’s results won’t come as a shock to you. He found that, for an older group of participants who had, on average, graduated from high school forty-eight years earlier, just 6 percent of the total classmates’ names were recalled. Intriguingly, however, he found that, for a group who had graduated quite recently (on average, three months earlier), performance wasn’t much better: just 15 percent of the classmates’ names were recalled.6

  At one level, this is deeply unsettling. After all, these are the people that we laughed and cried with during a very important par
t of our lives. But even more than that, the result just doesn’t feel right: surely we can remember more of our classmates than that. And that intuition is correct. When Bahrick asked his participants to simply recognize which of four names belonged to a high school classmate, performance was excellent. For those who had graduated three months earlier, name recognition was 90 percent. And recognition remained at very high levels (80–90 percent) for up to thirty-five years following graduation.

  The message to take away from this study is that the assessment of memory depends critically on how memory is measured. You really didn’t forget your classmate’s names: you simply became less able to spontaneously recall them. However, you can still pick out a classmate’s name when it is placed with others that are unfamiliar to you.

  Bahrick conducted a similar experiment on memory for high school Spanish.7 Even though participants in this study had a difficult time recalling a Spanish word when given only the English word as a prompt, they could still pick out the correct Spanish word on a multiple choice test. This result occurred for individuals who had studied Spanish up to thirty years prior to the experiment; however, it also depended on how much Spanish they had studied and how well they had performed in the courses. In other words, the longer someone had studied Spanish, and the better their grades had been, the more likely they were to retain the information.7

  To account for this impressive performance, Bahrick proposed the concept of permastore for memories that are extremely durable, resistant to forgetting, and that can last for more than twenty-five years. He suggested that memories acquired over a long period of time are more likely to end up in a permastore state. However, for information to end up in permastore, two conditions must be met.

  First, the information must be overlearned. For example, Bahrick asked participants to pick out the meaning of the Spanish word feliz. They had to choose from among the English words happy, fault, feet, new, or clean. Because feliz is such a common word in Spanish, students were probably exposed to it repeatedly. That is, feliz is likely to have been a Spanish word that was overlearned, and therefore, the meaning of feliz was less likely to have been forgotten. (Feliz means “happy” in case you were wondering.)

  Overlearning was first described by Ebbinghaus, and a large body of subsequent research has underscored its importance. So if you are trying to learn vocabulary terms and believe that you’ve memorized them, you would be well advised to continue reviewing these terms periodically even after you have the subjective impression that you know them. Overlearning will help you to put these words in your permastore.

  We have already briefly mentioned the other component that is necessary for permastore, which is distributed practice. Distributed practice means that exposure is spaced out over time. Returning to high school once again, you may recall neglecting the content of a particular course, and then staying up all night to cram for an exam. Although this sort of learning can be effective over the short term, and is better than not studying at all, it seldom leads to the formation of durable memories.8 For example, if you are going to spend ten hours studying a foreign language over the next week, you’d be much better off studying for ninety minutes a day rather than studying for five hours on two days. Although five hours spent studying over two days is better than not studying at all, it’s also not the best use of your time. When it comes to practice and retaining information over a long period of time, just remember: Slow and steady wins the race.

  Take It Personally

  Imagine that you’re taking part in a psychology experiment. You’ve been told to read and comprehend target words as they appear, one by one, on a computer screen. You’re also asked to answer a question about each of these words. Let’s say the word selfish appears on the screen, and you are asked “Is the word printed in capital letters?” Obedient participant that you are, you press the “yes” button. Other participants might see selfish and be asked, “Does the word rhyme with weight?” (in this case, no), or “Is the word a type of fish?” (again, no). Finally, some participants might be asked, “Does this word describe you?” After responding to a series of such word and question pairs, you are then given a task you weren’t expecting. You are asked to recognize as many of the target words as you can.

  Do you think the questions you were asked might have an effect on which words you were able to recall? If you have good memory for the previous discussion on levels of processing, the answer would be yes. As you might expect, questions about whether words were printed in capital letters led to relatively poor rates of recognition for those target words. Likewise, questions about whether two words rhymed also resulted in unimpressive performance. As we noted before, in both cases, this is to be expected, because these questions were answered using shallow processing. On the other hand, memory performance was better when participants had to think more deeply about the word (e.g., recognizing that selfish is not a type of fish). But the participants best remembered those words when they thought about the words in relation to themselves. This phenomenon is called the self-reference effect.9

  But is the self-reference effect really the result of being asked to consider words in relation just to oneself, or might a similar effect be obtained if the words were related to any person? Follow-up research used a self-reference condition like the one we just described, but also employed an “other-reference” condition. The idea was to have participants think about the words in relation to a familiar individual, but not someone they personally knew. For many of these studies, the go-to other person was Johnny Carson, the talk show host who had appeared on television every weeknight since the early 1960s. This was someone who would have been instantly recognizable to the participants, but probably not someone they knew very much about. (Carson, unlike many of today’s celebrities, was fairly circumspect.) When the researchers compared the participants’ recognition of the trait words in the two conditions, they found that words thought about in relation to oneself were better remembered than the words thought about in relation to Carson. In other words, the effect really is a self-reference effect, and not just a consequence of thinking about whether particular traits describe any one person.10

  The self-reference effect has been explained in several ways, and the arguments tend to be complementary rather than opposing. The first has to do with the emotional dimension of the trait words. It probably doesn’t matter to you, one way or another, whether Johnny Carson was selfish or incredibly generous. However, when thought about in relation to yourself, the question will engender a variety of emotions (perhaps pride about what a philanthropist you are, or guilty regret that you throw away all those appeals from charity that you receive in the mail).

  A second and perhaps more important explanation has to do with the nature of the self. You may now be struggling to learn a foreign language, but you can take some solace from the following fact: you are an expert on at least one subject, and that is the story of your life. In fact, you know more about yourself than anyone who has ever lived. Although this may seem a bit narcissistic, it is undoubtedly true, and it has important consequences for memory. When we’re asked if the word selfish describes us, we are able to think about the question by drawing upon specific autobiographical memories, such as last week when you treated your colleague to lunch (“See! I’m not selfish!”), or your unwillingness to buy cookies from the Girl Scout who appeared at your door last month (“Well, maybe I am, a little bit”). In the case of Johnny Carson, you probably aren’t able to recall specific episodes of generosity or stinginess. Researchers have described the self as “a well-developed and often-used construct.” In fact, the self-reference effect is so strong that we are more likely to remember the birthdays of others if those birthdays fall close to our own.11

  It’s possible to profit from the self-reference effect in a variety of ways. So the news here is good indeed. By attempting to learn a foreign language later in life, you can draw upon your extensive life experiences as rich retrieval cues for co
mmitting to memory the elements of a second language. Of course, not all concepts you need to remember lend themselves easily to self-reference. But thinking deep thoughts is much better overall than processing information shallowly. When you can use it, self-reference will be there as one of the arrows in your cognitive quiver.

  Emotional Aspects of Memory

  It probably comes as no surprise that the ability to learn a new language is affected by one’s mood and emotional state. Cognitive scientists have studied the links between memory, mood, and emotion. This section explores how to create an emotional environment that will facilitate pleasurable and effective foreign language learning.

  Think Positive

  In exploring the link between thinking and feeling, one basic tenet to keep in mind is that positive information is processed more efficiently and is remembered better and longer than negative information. The superiority of positive over negative information has been shown repeatedly in a wide range of studies, including ones that looked at memory for vocabulary words, grammatical structures, and the content of dialogue and text. This bias toward positive information may be of particular importance (and perhaps relief) to the adult learner because it has also been shown that unpleasant memories weaken over time. Margaret Matlin and David Stang called the overall trend for cognitive systems to favor positive information the Pollyanna principle, named after the little girl who in the popular books and movies focuses on the bright side of life in even the most depressing circumstances.12

  Of course, it would be impossible to achieve any level of mastery in a language by studying only positive vocabulary words or by producing only affirmative sentences. However, because positive linguistic features are easier to process, remember, and recall than negative ones, when it comes to doing things with the language—such as telling a story, making a presentation, or engaging in conversation—you will give yourself an advantage if you approach these activities from a positive standpoint. For example, it will be easier for you to create in your target language a sentence like “The president is a woman” than it will be to create the sentence “The president is not a man.” It will also be easier for your audience to understand what you are trying to say, since listeners also process positive information more easily than negative information.