Three cups of language

At a popular bar in Tainan (Taiwan), a Polish-Canadian girl was celebrating her graduation from the International MBA program. She spoke English at near native level, with only an occasional hint of her Polish origin. Two new IMBA students had been invited to the party; they had just arrived from Poland and were happy to have found one of the few people on the island who spoke their language.

The Polish-Canadian girl switched easily between Polish and English, speaking to various friends and well-wishers. At one point, she turned to say something to a Chinese colleague and I was shocked at how unbelievably bad her Chinese was after two years of living and studying in Taiwan. The syntax was off, the grammar wrong, the pronunciation flawed; I barely understood her. The native speaker she addressed didn't even realize she was speaking Chinese.

How could this be? This girl was obviously multi-lingual. She had just completed two years of studies in Taiwan. Why was her Chinese so bad?

To understand the dynamics at work, we must first dispel a common myth. Many people think that bilingual people are exceptional language learners, but this is not necessarily true. More on that in just a moment.

A Canadian friend present at the party asked, "But isn't it true that the first (foreign) language is difficult, the second easier, and the more languages you know, the easier new ones are to learn?" The question is a good one. Is learning a third language easier than learning a second? And would that mean that learning a fourth language is easier still?

The short answer is "yes" but this only applies to LEARNED languages. Your native tongue, no matter how many you may have, does not count as a learned language. Every person, in every culture, everywhere in the world inherits a native tongue. And if he lives in a culture with two or three official languages, he simply acquires two or three languages, regardless of intelligence or language learning ability. In China even small children can speak Chinese. Does that mean that they are more intelligent than you because you cannot?

Nearly 70% of Taiwanese grow up speaking both Mandarin and Taiwanese. Many also speak Hakka or some tribal language. Since they grow up with two or three languages, one would assume they are good language learners. But even after many years and countless hours of English classes, they generally score in the lowest percentiles on competitive exams when compared to other non-English speaking countries.

There was a Chinese-Canadian girl studying with me in Costa Rica. She was absolutely incapable of stringing a decent Spanish sentence together after a year of studying in the country. Before arriving in Costa Rica, she completed two years of Spanish at university and was given perfect marks.

One night we went to a Chinese restaurant in San Jose. The Chinese-Canadian found that the waiter couldn't understand her English. She asked me to translate into Spanish, which led to the discovery that the waiter's Spanish was less than basic. "Couldn't you just talk to him in Chinese?" I asked. This was long before I spoke any Chinese.

"Oh, no, I could never do that." She responded. "Chinese is a home language. I can only speak it with my family."

Your native tongue doesn't count as a learned language because you neither studied it nor learned it in an academic sense. Your mother didn't force you to memorize lists of irregular verbs. She never made you diagram sentences or do dictionary exercises. But when you study a language in a traditional program, you will be asked to do all that and much more.

When I was studying applied linguistics at the University of Main (Germany), we had any number of bilingual, trilingual and multilingual people coming there to study. At first most students envied these "linguistic geniuses." We marveled at the ease with which they switched from French or Russian to German.

When exam time came, however, we discovered that bilingual students were no better at doing homework, completing assignments or memorizing grammar and vocabulary. In fact, many were worse. They had spoken the language for so long, often incorrectly, with their families that mistakes and shortcomings were cemented in place.

I remember a Mexican-American student telling our Spanish teacher a story about what happened to him while walking along the train tracks. He referred to the train tracks as las cosas para el tren, the things for the train. The standard term had actually been on a recent exam, but he continued to talk the way he had when he was five years old and didn't possess sufficient vocabulary to express himself fully.

This brings up another point about bilingual people. Although many claim to be bilingual, it doesn't mean that they are 100% fluent in either or equally fluent in both languages. We acquire our mother tongue from our mother, hence the name. Most of this acquisition occurs before we attend school. For monolingual children, school reinforces what they have already learned at home. For bilingual children, however, going off to school may signal the end of the development of their "home language." Vocabulary and usage may become frozen at the level of development of a five or six year old. The Mexican-American kid will go on to take course in biology, chemistry, history, and literature, all taught in English. If he is clever, he will eventually go on to university. Maybe he will become a lawyer or an engineer. Will he know all of this specialised vocabulary in Spanish? Monolingual kids acquire this specialised vocabulary in school. Bilingual children usually don't acquire it at all in one of their languages.

Before I'm accused of being anti-bilingual, let me say that I have also known brilliant linguists and translators who were raised bilingually. But they became brilliant by attending school. The only advantage they had over classmates was that their pronunciation and accent were usually better.

In my case, I grew up constantly exposed to Spanish and Italian, but didn't actually learn the languages until I went off to college. When I worked in the financial industry in New York, I knew Spanish vocabulary for accounting and finance only because I had attended business school in Costa Rica. Said another way, when I was five my grandmother never taught me the words for exotic options, hedge strategies, or tax avoidance. Grandma and I never discussed the merits of covering a stock position with puts and calls.

So, whether you agree or not, it is clear where I stand on the subject of bilinguals acquiring a third language. Now let's deal with the other part of the theory. If you have learned two languages through study, is a third easier? I'd say yes, but.... We need to understand how language is acquired, and even experts are divided on the subject. 

David Long, head of the Thai language program where I studied in Bangkok, explained his view on language acquisition like this:  Picture an empty cup in your brain. You fill the cup by listening to a language. When you have listened enough, the cup will run over. The overflow is speech. In other words, you listen first and then you speak. Sufficient listening leads to speaking correctly.

David's illustration is drawn from a language acquisition theory called ALG (Automatic Language Growth), which maintains that students must listen for an incredibly long time before being allowed to speak. Other theories, such as The Silent Way, also require months of listening before speaking, but ALG is one of the few theories that quantify how much listening it takes to fill the cup. The number of hours varies depending on the language being studied and one's mother tongue, but for most people, one to two thousand hours of listening are required before speaking can commence.

ALG and other theories look at the way a native tongue is acquired. We have already established that the native language is acquired by first listening, mostly to one's mother. How long do babies listen before they speak? Most children don't start speaking till somewhere between two and three years old. Some may possess a vocabulary of between five and twenty words by age two. Babies listen for a long, long time before uttering a word.

The reason why most people find learning a language difficult is because they don't get enough input before they try to produce output. While listening is the best way to acquire language, the second best way is reading and studying. Whatever the means of input, you need to hit your thousand hours mark.

An American friend, also in the IMBA program in Taiwan, recently said, "After nearly four years in Taiwan, I am finally ready to admit that I am not going to learn the language by osmosis. I signed up for classes yesterday."

The IMBA, like most such programs, is taught in English. Foreign students are not exposed to Chinese in the classroom. They are also not exposed to Chinese while they are studying or doing homework. Most students support their studies by working as English teachers, another activity that does not provide exposure to Chinese.

David Long is really big on laying out time charts of daily routines to demonstrate just how little local language exposure the average expat receives.

07:00 - 08:00:             Wake up, shower, eat breakfast in the room while watching CNN.
09:00 -12:00:              IMBA classes, in English
12:00 -15:00:              Lunch, nap, homework, study, meet with English-speaking friends to complain about Taiwan.
15:00 -19:30:              Teach English.
19:30 -20:30:              Eat dinner in the room watching illegally-downloaded episodes of American TV shows.
20:30 - 00:00:             Study, homework, gym, beer, complain about life in Taiwan.
00:00 - 07:00:             Sleep.

It's possible for the average foreigner to get through a whole day without uttering a single word of Chinese (except to order food). We could be generous and say that people with similar schedules are exposed to 20 minutes of language per day. But it's not new language; it's the same 20 minutes they had yesterday: "What would you like to eat? Is that for here or to go? Would you like to take advantage of our new buy-five, get-three-free promotion?"

That twenty-minutes is probably worth about five minutes in terms of learning. At five minutes per day, how many days do we need to reach 1,000 hours?

In actuality, at this rate you could never learn a language. At five minutes per day, the rate at which we lose language surpasses the acquisition rate. In fact, if your exposure is less than a solid hour per day, you will probably never learn a language.

Think about one of your language lessons. You learned a number of words and phrases; perhaps you even completed some exercises in class. Ten minutes after class you would have forgotten fifty percent of new vocabulary. By the time you sat down to do your homework, it was almost like seeing a brand new list of words. That's why significant, repeated exposure and review is necessary. If you don't do your homework, that day's lesson may be completely lost.

A good example of loss: all the Americans who suffered through four years of high school Spanish, but when they went to Puerto Vallarta on holiday, discovered they couldn't say anything beyond, "Donde está la biblioteca?" With their level of Spanish, the library wouldn't do them a lot of good anyway.

Let's say that you have successfully acquired a foreign language after two thousand hours of exposure. Could you acquire the next language in less time? 

ALG says that when you start to acquire a second language you simply establish a second cup in your brain. A portion of your first cup was simple language mechanics that have to be mastered before any language can be learned. But if you have already learned them once, they're there. In simple terms, your second cup would already be 15% full. So a third language might be marginally easier, but for the most part the 3rd, 4th and 87th languages would be equally difficult to learn.

Keep in mind that we are talking abut language acquisition, in general, not about the difficulties in learning specific languages. Related languages, of course, would be easier to learn. For example, if your second language is Khmer, you will acquire Thai faster than someone whose second language is German. Learning Spanish is easier of you already speak Italian.

Although David Long and I would both agree that this is true, we vary slightly in why we believe that a related language is easier to learn. We concur that there are linguistic triggers; for example, the smell of Thai food triggers Thai language in my brain. David says that the reason why a Khmer speaker learns Thai faster is because cultural understanding is one of the most important aspects in language learning. I agree, but I still can't get away from the nuts and bolts; on some level I'm a language mechanic. I hold that a related language may be made easier by similarities in grammar and vocabulary. ALG always says don't get hung up on words; language is about communication. I agree, but I still like to see my students copying vocabulary lists.

It's a minute point which language geeks like myself enjoy debating. But the practical point we agree on is that it is slightly easier to learn a language related to one that we already know.

On a personal note, I picture boxes in my head labeled with the names of languages I speak to some degree. There are boxes for English, Chinese, Thai, Khmer. I studied some Filipino and learned a bit of the language when I was at school there, but not enough for it to have a separate box.

So, what happened to the Polish girl? Here's what I think and it goes along with the above-mentioned boxes.

We all have a box in our brain labeled native language (e.g. English). Many people have studied some foreign language (e.g. French), but most don't progress to any level of fluency. This means that a box marked French never appears in their brain; in its place they get a box named foreign language. Into that box will be thrown all the remnants of high school French, college Spanish and that six-week Berlitz course in Japanese.

Fellow students in Thailand often told me things like this: "Every time I try to speak Thai, my high school Spanish comes out of my mouth. And I don't even speak Spanish." Right - it's that foreign language box speaking. After 500 hours of study, Thai will have its own box and this type of interference will stop.

In the case of the Polish girl, I believe that as a baby she acquired Polish as her native tongue. Her family then moved to Canada before she had finished elementary school. At first, any English went into the foreign language box; perhaps she responded in Russian or German when someone asked, "What's your name?" As she attended school conducted in English, that language eventually got its own box. Once that happened, the only interference came from Polish. 

After a number of years in Canada, there came a time when her total exposure to English exceeded that to Polish. In fact, she didn't even possess Polish vocabulary for half of what she was learning in school. At that point, English jumped up a notch and became the de facto mother tongue, and Polish became a second, extremely fluid, foreign language. Along the way, her high school French lessons were shoved into a box with the German and Russian she had learned at school in Poland. She was probably no better, and possibly worse, at learning French than her mono or bilingual Canadian classmates.

When she came to Taiwan for the IMBA, she took Chinese classes because she had an interest in learning languages. The Chinese lessons went not into her English box, but into the next one in the hierarchy of her brain - Polish.

I noted that she spoke Chinese with a purely Polish accent, uninfluenced by English. If we had a native Polish speaker to help us analyze her Chinese syntax and grammar, I bet we would find that most of her mistakes were typical of a Polish rather than an English speaker.

Amazingly, she no longer makes mistakes in English. This would support my thesis that English has become her dominant tongue. But it is her original tongue, now her strongest foreign language, which interferes with her acquisition of Chinese.

In theory, no matter what your native tongue is or what language is producing interference, you should be able to learn Chinese after 2,000 hours of listening. But being in the IMBA program and not a Chinese language program, she had probably had less than a few hundred hours of Chinese input. Most IMBA students who study Chinese on the side don't put too much into their classes because they are not part of the degree program. And they often drop out of their Chinese lessons in order to concentrate on the IMBA. At that point, their level of exposure to Chinese is the same as any other foreigner in Taiwan - nearly zero.

On a psycholinguistic level, I wonder if she may even have an emotional attachment to the language she spoke as a young girl. Perhaps on a subconscious level she doesn't want to let go of it. Could she be so emotionally attached to Polish that she will never learn Chinese? I don't know. I write, and throw my ideas out there in the hopes that readers will write back and give me some input.


Antonio Graceffo ( lives in Asia and writes on martial arts and adventure travel.  His website is

Recommended citation format:
Antonio GRACEFFO. "Three cups of language". December 8, 2008. Accessed October 19, 2019. <>.