Why don’t immigrants’ kids properly acquire their parents’ language?

My colleague Rome Lacvrencic, head of the B.C. Association of Teachers of Modern Languages, and I had an interesting Twitter discussion recently. 

Lavrencic, of Polish extraction, heard some Polish at home in Ontario, Canada, English everywhere, and was in late French Immersion. By the end of Grade 12, he says he was “more proficient in L3 than L1.”  He attributes this to being able to speak more French than Polish. 

 This  is a familiar refrain: “I used to be good at ____ but now I don’t speak it much so I’m bad at it.”

This was where I disagreed. I told him that speaking wasn’t the point, but that listening was.  

So I thought I’d take a look at this via numbers and my own experiences. 

My L1 was German.  I heard it at home a lot until Grade One, and much less after Grade Four, when my cousin Sig came to live with us.  Sig spoke Spanish, French and English, so English it was at home. 

Now, when I speak German, I sound like a five-year-old from 1963. I hear my folks speak German but that’s about my only exposure. And I suck at German. When I am around German speakers, I understand a ton but I can say much less than I understand. 

In terms of input, mine dropped to close to zero at age 9. Lavrencic went through a roughly similar process: Polish dropped off but French input massively upped.  My guess is that he (and anyone else in his shoes) would get 5-6 hours daily of French input at school, plus homework (reading) while in Polish (like me in German) would have gotten maybe an hour or two.  

Lavrencic took French in Uni and also teaches it so he’s obviously super-proficient.  

In my view, Lavrencic is bringing up the problem of post hoc, ergo propter hoc (after this, therefore because of this), also known as the correlation vs causation problem. There was speaking and input, then there was acquisition.  The acquisition happened after both the speaking and input. Was it therefore because of the speaking? 

The research says it’s the input. Terry Waltz recently remarked, echoing Krashen, that there are loads of cases of people acquiring languages without speaking them. The deaf who do not get speech training are one. As we all know, when we start acquiring a  language, we go through Krashen’s “silent period” where we understand more and more but our speaking lags.  It is also well-known that babies as young as a few days have begun understanding some aspects of language 8 months prior to even single words emerging. 

Recently on Yahoo this topic came up and master teacher Hai Yun Lu weighed in. She’s Chinese, married an American, and wants her kid to acquire Chinese. Check it:

“I have raised my son to be bilingual. There are many rules and  practices we have implemented at home in order for this to happen. After my son was born, a college professor visited me and shared research she had read. If I wanted to raise a bilingual child, then his second language input needed to be minimal 30% of his total language input (I wish I could find this actual research to share with everyone).

Let’s say, if his waking/alert time is 14 hours a day. 8-9 hours in daycare = English input. He has about 5 hours at home with us. Listening to me speaking Chinese to him, his father speaking English with him and his parents conversing in English. Of course, on the weekends/holidays, he gets more Chinese input. Still, we can barely meet the minimal input amount. Therefore, rules have come into place in our house. Each time we go back to visit China, first and most, we carry a suitcase full children books back for him. (Richard Scary’s collections, Curious Gorge, Clifford…) I only read to him in Chinese, even with an English book [she means, she reads the words to herself silently in English but says them in Chinese].

We rarely turn on TV before he goes to bed. If he’s interested in watching some cartoons, I do whatever I can to get them in Chinese. Therefore, he watches his favorite cartoons in Chinese (e.g Thomas and Friends, Disney films, Curious Gorge, Magic Flute’s Adventure). The majority of his playmates have been Chinese-speaking kids until this spring. He has developed close friendships in JK, where we have finally “extended” our friends circle.

My son is one of the very few kids who can speak Chinese fluently, in comparison to the kids in a similar situation. Many people complain to me that their kids understand their languages, but only speak back in English. I always say “input” proceeds “output”. They need more comprehensible input before they can output. (Here I have left out some psychological factors such as the desire to “fit in”, which typically occurs once when kids start school and they start to refuse to speak their parents’ languages.)

Many of my son’s friends’ parents are very eager to have their children to speak Chinese, and they keep saying to me: “just speak Chinese to my child, I hope we will be able to speak.” It hasn’t worked for any of his friends yet, because what we can say to each other is incomprehensible to his friends, unless I want to turn a playdate into a Chinese lesson time.”

Haiyun Lu

What Is Personalisation? Two Approaches.

What is “personalisation”?  We all agree it matters.  My definition: personalisation is any connection between subject matter and individuals’ interests and characteristics.

A very talented District colleague recently did me the favour of Twarguing with me.  She posted a picture of a bicycle with some Spanish sentences explaining the value of riding a bike, thus: 

For the non-Spanish-speakers, the sentences include “puts a big smile on your face” and “reduces the risks of heart attack,” etc.  These are all about the advantages of riding a bicycle regularly.  I would never use this with kids, cos like O.M.G. it’s boring, LULZ but anyway.

When I saw this, I looked at a few words from the picture Wiktionary’s Spanish frequency lists.  Most of that picture is low-frequency vocabulary (i.e. not in the top 1500 most-used Spanish words).  So, I responded with the following question:

How is low-frequency vocabulary “important”? If your reference is to grammar (e.g. 3rd person verbs), this better taught w/ high-frequency vocab.”


Shauna here is suggesting that students investigate their own interests and use language pertaining thereto.  In other areas, we have suggestions about using project based learning and genius hour in the language classroom (with excellent rebuttals (especially for genius hour proposals) from Sarah E. Cottrell).

We are here getting into a classic traditional-methods-vs-comprehensible-input teachers’ argument:  do we make language class interesting– and personalised– for students by

a) recycling high-frequency vocabulary, or

b) by allowing students to choose their own vocabulary for activities?

A general note: while we all take some interest in what others do/like etc, there are limits.  Walk into a Joshua Tree fire circle, and if you’re not a rock-climber, you’ll be baffled and bored within minutes, because “it’s slammer left-facing hands on 2s to a sidepull and then a mantle over a crappy blue Alien and a slab runout” is basically irrelevant to you.  In any social situation, there is a balancing act between interest in others’ stuff and being bored.  So it is in a classroom.

Textbook personalisation suggestions have a number of basic problems, one of which is keeping kids interested.  Why should Johnny want to listen to/read the vocab about ordering dinner, or recycling, or bargaining for fruit in a French market, over and over?  It’s not that these activities and words are boring per se, but when was the last time you spent three weeks using forty words and one grammar device to discuss the same topic? Never– because that’s boring. So the simple answer– for teachers who do not use stories– is, let the kids pick and choose their vocab.   T.P.R.S. and A.I.M. teachers, as we shall see, don’t have many challenges keeping kids interested.

But if students choose their own vocab for class activities or projects, there are five big problems.

First, there is the problem of usable frequency.  If we want to build functional fluency in any language, our first priority is make sure students acquire the most-used words before the less-used.  Obviously, there will be exceptions: “communicative” teachers typically like to make sure kids know all the words for school things such as pencil, desk etc, while we comprehensible input people like animals etc for our stories.  Now, a student may be into activities that use high-frequency words.  But much more often, the opposite is the case.

If Johnny is into, say, bicycle racing, and Sheila likes wrestling, great.  But how often is Johnny going to hear/read cycling-related– and Sheila wrestling-related– words?  The answer:  in most language communities, especially ones to which people in their first five years of language acquisition belong, not very often.  This means they are putting effort into something which has limited communicative value for them and for others in their class.

Second, we have the problem of shared interest.  As I noted above, if Baninder likes Call of Duty and Maricela likes chess, what– as relative beginners– are they going to talk about?  Maricela is probably not going to be especially interested in hearing about shooting people, team missions, ammunition etc, and Baninder is not going to want to hear about endgame strategies and Sicilian openings, etc.

In the “real” world (probably online), Baninder can find his own C.o.D. crew in French and Maricela can play chess with French speakers, but in class– where realistically 95% of language acquisition happens for our students– how are we going to get each kid– not to mention the rest of the class– to “buy into” hearing and reading others’ specialist vocab?

(As an English teacher, my first great reading realisation years ago came from my brilliant colleague Louise Hazemi, who in Surrey pioneered the use of literature circles for novel reading.  We used to have a “novel a year” system, where kids were assigned To Kill A Mockingbird in 10th grade, Lord of the Flies in 11th, etc.  The problem?  Most kids hated these books (either because they were “too hard,” or simply because they had been assigned), didn’t read them, cheated on tests and essays, etc.  So, at our school, we asked the kids what books they would like (and asked teachers) and for each grade bought 10 copies of 8 novels.  Now, the kids in each grade pick a novel to read (yes; we still offer Mockingbird and L.o.t.F.) and BOOM! all the kids read at least one novel, and all the kids report enjoying their reading (they still do essays, discussions etc about their chosen novel).

It is much the same with silent reading.  I start each English and Social Justice class with 20 minutes of silent reading.  There are three silent reading rules:

  1. You must read a book (no newspapers or magazines) and not talk, listen to music, or use your phone.
  2. You must not read anything from any class during silent reading.
  3. Your book must not suck.  If it does, get another one.

How does it work?  Brilliantly.  While my less-literate boys grumble at the start of the year, after a week every kid reads and every kid likes reading.  Probably two-thirds of kids read young adult novels, while another third prefer things like biographies, how-to books, various factual genres, self-help, etc.  This is because they choose things that are interesting (and readable) to them and because there is no “accountability piece.”  No “book report” marks, reading logs, etc.  As long as they are reading and enjoying their reading, I am happy.

Now at this point I can see Madame Nero (and any other person who shares her view about how to personalise the language class) saying “Exactly!  Let language kids do the same thing! Let them decide!” However, the key here is that nobody is forcing the kids to learn/acquire things which they are not interested in.  Kids like free reading because it’s free:  they aren’t forced into something they don’t care about.)

Third, there is the quantity of input problem.  We know that what people acquire is a function of how much comprehensible input they get.  They need to hear the words or structures a lot to first recognise them automatically, and even more to be able to automatically say them.  So if we are going to run our class around student-identified student interests, how do we deliver 30 different sets of vocab often enough that the kids– even if they want to, which we are not guaranteed– pick them up?

Say each kid gets to decide 5 words germane to their interests which they want to have incorporated into class activities.  That’s 150 words.  That’s half of a year’s recommended vocab load right there!  As we very well know, it’s simple math:  the greater a variety of vocab we use, the less time we have on each word which means poorer acquisition of each word. As the great Terry Waltz recently noted, if you want the kids to acquire more words, teach fewer words. There is also the challenge of integrating specialist vocabulary into teacher-planned activity.

Fourth, we have the output problem.  In many traditional classes, it is assumed– wrongly– that if kids “learn” vocab (and grammar) and present it in some way, they are picking it up.  This is simply wrong, as the research shows.  And, learners by definition generate error-filled and impoverished (two-dimensional) output.  I do not see the point of making other learners listen to that.  As Terry Waltz has famously said– with Stephen Krashen agreeing– “peer to peer communication is the McDonalds of language teaching.”

This is what is supposed to happen in a “communicative classroom (here, two Vietnamese speakers are learning English):

Thanh: Where Michael today? He here?

Vien: Where is Michael today?  He is not here.

Thanh: Ahh, yes, where is Michael today?  He is not here.

Here, Vien– who is also learning English– is supposed to notice Thanh’s error, “remodel it” properly, so Thanh can fix his output.  Now, here is what would actually happen:

Thanh: Where Michael today?  He here?

Vien: Michael home.  He not here.

Thanh: Ah, yes, Michael home.

Even though Vien and Thanh want to learn English, and are working away at it, they will inevitably produce poor output (for a variety of reasons).  So the ideal situation described above generally does not happen with two learners.  If Thanh’s interlocutor was a native (or very competent) English speaker, this communicative activity would probably work.

Fifth, there is the dictionary/Internet problem.   As soon as the kids want to generate their own vocab, we know what they do:  they fire it into Google translate, and we know the results.  So it becomes the teacher’s job to edit word lists, activities, presentations, rehearsed dialogues, etc.  I don’t know about you, but that’s boring and often I am myself scrambling to figure out how to say _____ in Spanish.

So, if our goal is to deliver a ton of compelling and multidimensional high-frequency language, and to repeat that language over and over so students hear it often enough that it gets wired in, the “choose your own topic” idea won’t work.   But the question remains, how do we personalise vocabulary and maintain student interest? 

One answer involves using the world’s oldest and most-proven teaching method;  stories.  Everybody likes a story, because we naturally find people and their hopes, problems etc more interesting than things or ideas, and because suspense– what happens to ____?— is another universal hook.  Stories are always more interesting than any other kind of input.  Everyone can relate to basic human questions such as wanting to have ___, being scared of ____, liking/disliking someone, etc.

In T.P.R.S., our use of parallel characters and parallel problems allows us tremendous room for personalisation.  If we’re working on esperaba que ____ le explicara… (“s/he hoped that _____ would explain…”), say, I can have a boy who wants to have the mysteries of talking to girls explained to him (great topic for all teens: boys want info, girls will think it’s hilarious) and a girl who wants to have say Call of Duty explained to her.  (Stereotypes are great to play around with).  This works even better when we know students, and we can throw a kid (and their interests, from say our start-of-year questionnaire) into a story.   If we know Breleigh likes dogs, hates cats and looooves Ashton Kutcher, well, Breleigh eseperaba que Ashton Kutcher le explicara por qué no le gustaban los gatos.  Any half-decent storyteller can get the audience to empathise with or at least be interested in a character who is a bit different than they are.

Another answer involves recycling high-frequency vocabulary in a way that ackowledges student interests and preferences.  For a rank beginner, something like owning a specific kind of pet, or liking or disliking any kind of thing or activity is a great start.  In my first story, Los Gatos Azules, a boy wants to own ten blue cats.  So, we personalise by asking the students the same questions we ask our actors:

Here’s an example from my Level 1 class:

Me: Ace, ¿tienes un perro?  Do you have a dog?

Ace: No.

Me: ¿Tienes un gato? Do you have a cat?

Ace: No tengo gato. I don’t have a cat.

Me: ¿Te gustan los gatos o los perros? A mí (pointing at myself), me gustan MUCHO los gatos. Do you like cats or dogs?  Me, I REALLY like cats.

Ace: Me gustan los gatos. I like cats.

Me: Clase, levanta las manos si te gustan los gatos. (half of class raises hands, so I point at a kid who didn’t). Mandeep, ¿te gustan los gatos? Mandeep, do you like cats?

Mandeep: No.

Me: What did I just ask you?

Mandeep: Do you like cats?

Me:  Mandeep, los gatos– ¿Son simpáticos, o no son simpáticos? Cats– are they nice, or not nice?

Mandeep: No.

Me:  ¿Los gatos no son simpáticos? Cats aren’t nice?

Mandeep:  No.

Me: Class, what did Mandeep just say?

Class: Cats aren’t nice.

So, here we have some personalisation: the kids are explaining their opinion about cats and dogs.  This is basic stuff.  (Note:  I am not expecting any output other than y/n here (though if the students want to say more, they can).  My only aims are that they understand what is being said and that they can connect the vocab to their selves or interests.

Here is a level 2 example of personalisation. In the story we are doing, a Dad is chewing his kids out for not having done homework and chores.  So we are acquiring what did you do? and I prefered, etc.  In the story, Dad asks his kid “What did you do last night?” and she says “I went to Cabo San Lucas and talked for 9 hours with Dave Franco.”  Dad asks “Did you do your homework?” and she says “No, I didn’t, Dave did it.”

All we have to do in P.Q.A.– personalised questions and answers– is ask kids in class the same questions we ask our actors.

Me: Breleigh, ¿que hiciste anoche?  What did you do last night?

Breleigh: No hice nada porque tenía que estudiar. I didn’t do anything cos I had to study

M: ¿Qué estudiaste anoche? What did you study last night?

B: Estudié la biología. I studied bio.

M: ¿Qué querías hacer anoche: estudiar, o bailar? What did you want to do last night: study, or dance?

B: Quería bailar. I wanted to dance.

M: John, ¿qué prefieres hacer tú– bailar, o jugar Call of Duty?  What do you prefer to do: dance, or play C.o.D.?

J: Prefiero jugar C.o.D. porque es más interesante. I prefer to play C.o.D. cos it’s more interesting.

We can get an immense amount of mileage out of a fairly limited range of vocab, as you can see.  If we throw in some weird stuff, we can get a zillion more miles.  For example, I could ask Breleigh if she likes elephants (free cognate) more than cats.  If she says yes, we’re off: do you own an elephant?  what is a good name for an elephant?  etc etc. These details can serve in stories, and they are great for random “review” P.Q.A.

Now, these are simple examples, and I hope you’re seeing the point:  we can personalise without getting into specialist vocab.  Not every kid is into Call of Duty (or chess, or ballet, or gangster rap, or Peruvian food, or French culture), but a teacher who is willing to listen to kids will figure out what people have opinions about and get them to express those.

The teacher’s job in part is to explore student interests, but also to make the language classroom functional (comprehensible and interesting) for everyone, so sometimes you have to say “sorry, Johnny, that’s too complicated” or “nobody else is interested in that, sorry.”

Personalisation: people basically want their interests and selves acknowledged.  If Johnny says “I prefer C.o.D. to ballet” and Suzy “God, I hate cats,” that is good personalisation.  We acknowledge their interests and views, and we give them what they need: an ocean of repetition on limited vocab, varied by context, cognates and sometimes wacky fun stuff.

So, in a nutshell, to personalise properly:

  • avoid having students generate lists of words
  • avoid making students listen to/read low-frequency specialist vocabulary
  • connect students with high-frequency vocab by soliciting their opinions, or info about them
  • use stories and ask students the same questions as you ask actors
  • integrate students– or info about their real (or imagined) selves– into stories

NOTE: Teachers in an immersion environment are going to be able to use more vocab than the rest of us, and there is therefore going to be more finely-tuned personalisation, and at serior Immersion levels there will be way more room for vocab personalisation.  But for most of us…keep it simple is the way to go.

Why do T.P.R.S. teachers overpractice everything?

One feature of a T.P.R.S. classroom– especially during story-asking– is an apparently insane amount of repetition. The teacher is constantly asking questions such as “is there a boy?” and “are the boys dancing in the rain?”. We also use parallel characters to get even more repetitions on vocab.

Now, we know that the more we hear something that we understand, the more we acquire it. But, as it turns out, there is another very good reason to “overpractice” things.

In research done in Colorado, scientists basically did two things while watching subjects learn new skills (here, using a mechanical arm to move objects around). First, they showed that as people practiced more, they used progressively less energy to get a task done. Mastery was defined as, they could do the task with way less energy than they started (i.e. they got more efficient), they could correctly do it, and they could do it quickly and without thinking (little conscious focus).

Secondly– and more interestingly– they showed that once subjects had mastered the task but kept on doing it, their energy expenditure for doing the task kept dropping, while the subjects were unaware of this process. In other words, the subjects got better at the task without realising they were getting better.

When I told some professional musician friends about this, one guy immediately said “amateurs practice until they get it right. Professionals practice until they can’t get it wrong.

The implications for the languages classroom are enormous and I’m only starting to think about this. (Of course, Mr Intuition himself, Blaine Ray, figured this out twenty-five years ago– now science backs Papa Blaine up on it):

A) “overmastery” should be the goal. We want to make it so that first, students understand without any thinking or pauses. And second, we want them to be able to quickly and easily and unselfconsciously and fluently “spit out” responses to questions, or statements describing things, or questions, orally or in writing.

B) the reason we want this, as the article shows, is that once we have “overmastery,” we are using both much less mental bandwidth and less energy to process the “known” (acquired) stuff, which, crucially, frees mental processing power for newer input.

This last point’s importance cannot be overstated. You want as much “open mind” as possible available for what you are now learning. The more mental clutter you have– unconscious self-talk along the lines of wait, what does that mean? What did he say? How do I say ____ again?— the less energy and mental room you have for new stuff.

C) if you think mental processing power is unlimited, read The User Illusion. Norestranders shows that since about 98% of mental processing is unconscious, we have very limited opportunities for conscious input, and it follows from this that the more we limit the variety of input, and “smooth out” the mind’s unconscious operations– by “automatising” them with practice– the easier it will be to acqure new stuff.

It occured to me that this could potantially be a “communicative” teacher’s argument for yet more student talk in the languages class–“the research obviously says they have to practice”– but this is probably not the case. The reason for this is that practice, in the languages classroom, is literally 90%+ listening and reading. The more we listen and read, the more automatic the language becomes. Blaine Ray was asked at a workshop whether the actors in stories acquired more than the rest of class (most T.P.R.S. teachers seem to have a regular crew of actors) and he said “no.” You don’t learn (much) by talking; you learn a ton by listening.

D) If we “overmaster” the seven basic “power verbs,” we can do pretty much anything even with limited vocab. Has, is, wants, needs, goes, likes/loves, gives/receives I am betting will get 80% of the needed work done in any language. So if students have these automatised, they have some really good “real world” prep (a point made by Jim Tripp on Ben’s).

Musical analogy:

When I started with Irish music, every new tune was major labor. But I went to sessions weekly and after awhile they became faster and faster to learn. (My first-year goal was two tunes/week and I almost did it– I’d hear something awesome, ask its name, then look the music up online and practice at home). The learning became quicker because a) my fingers got the moves wired and b) I started to internalise the scales and rhythms. Irish is mostly in mixolydian and dorian, so the classically-trained and rock guy (me) will naturally play a C# instead of C natural in a mix tune. Eventually I just naturally unthinkingly started hitting the dimished 7ths of the mix and dor scales

Once you play (and listen) enough, the input soaks in and you stop making certain kinds of mistakes. You also subconsciously note tune and bar similarities and you find yourself going “well the B part of Scatter the Mud is more or less like The Noonday Feast except with a mix 3rd” or whatever.

And when these things start happening– when finger movement and scales (what’s “in” and what’s “out”) get automatised– acquisition of tunes becomes easier and quicker. Indeed, I now mostly learn purely by listening, and only occasionally look at sheet music.

Another interesting effect: the more “wired in”– automatised– tunes become, the more you can experiment on them. I still remember playing Cooley’s, a tune I’d played a hundred times at least, and finding myself playing triplets– hard on mando– because I had the tune so wired in that my brain could now explore a new music trick.

I also remember the exact moment I acquired the Spanish subjunctive. After a month of lessons, lots of travel and work, on a sunny day waiting for a bus near Huehuetenango, Guatemala, a guy asked me “when is the bus coming” and I said “espero que venga pronto.”. If the input is there and you get it, the more of the basics you unthinkingly do, the more easily the non-English stuff comes.

Anyway, the analogy’s point: the more you solidify the basics, the easier anything subsequent becomes. Luckily for us, T.P.R.S. allows us zillions of reps because stories– especially with multiple parallel characters– accomodate repetition without boredom because the language is the container, not the content: we teach stories with language, not language with grammar.