Questions and Answers
About
Teaching English to Non-English Speakers
The teaching methods that we specify in the video series Teaching English Pronuncation and Direct Instruction Spoken English are based on facts of student performance. These procedures are at odds with much of the literature on how to teach spoken English—in the amount and type of practice we provide, in the skill sequence that we follow, and in the specific techniques that we use for teaching sounds and words. These differences may lead you to question the thinking behind some of the details. If you have questions about our approach or if you don’t agree with us, email me at zig@nifdi.org and I’ll try to provide an answer. If you don’t mind, I’ll post the question and answer on this site.
Submit your questions to zig@nifdi.org
Q1-/r/ sound | Q2-final /s/ | Q3-/ts/ /h/ sound | Q4-Spanish Speakers | Q5-Deaf Students | Q6-Primary Language |
Question 1 comes from Roberta Kleister:
Your suggested procedure for teaching the /r/ sound is completely at odds with knowledge of the speech-production literature and accepted practices for teaching phonemes.
You indicate that non-english speakers who can’t produce the sound should be taught to smile broadly, then produce a growling /r/ sound.
In fact speech analysis documents that native speakers do not produce the sound in this manner. They purse their lips, as one does when whistling. Films of speakers clearly show this mouth-lip orientation. How do you justify your disregard for this normative fact?
Answer: Speakers who don’t have the /r/ sound in their language can’t produce it or hear it clearly, so they reduce it to something else, often a vague vowel sound. If you require them to purse their lips, you increase the possibility that they will produce the /w/ sound, as in wabbit. And it is VERY hard to correct them because they can’t clearly hear the difference between the sound they produce and the one you model, particularly since they are faithfully producing the lip-pursing component.
Here’s a simple exercise that demonsrates why smiling is an important part of the initial teaching.
Purse your lips and say the word rich. No problem.
Purse your lips and say the word which. No problem.
Smile broadly and say the word rich. No problem.
Smile broadly and say the word which. Big problem.
You can’t really do it, and we can shape the approximation you do produce quite easily by having you make it a clear growl, rrrrr.
So if you direct students to smile when learning to produce the /r/ sound, you rule out some of the mistakes they will otherwise make. Once they learn to produce the sound with this exaggerated routine, more natural behavior is shaped by the different sound contexts for the /r/ sound. For instance, words like or or you’re have a very difficult transition from the vowel sound to the /r/ sound if you try to smile for the /r/ sound, but the transition is not hard if you don’t smile. So these sound contexts are teachers for shaping the “natural” mouth-lip orientations.
That’s the long answer. The short answer is that facts about normal speech don’t provide any real information except what the goal of the instruction is, not about how to get there. Furthermore, the goal is pretty obvious without elaborate information about tongue and lip positions. If the tongue position is important as an instructional variable, it is quite visible, like the tongue position for /th/ or the /m/ sound. If tongue position is more obscure, the shaping is based on the responses the students produce.
Question 2 comes from Pam Corbyn:
I’m teaching a Thai student, aged 14.
A common pronunciation error (or one of them anyway) is dropping /s/ from the end of words.
There are other problems too with some adjacent consonants but the dropped /s/ changes the meaning.
Answer: An effective way to go at it is to:
1. Stop the student right after the mistake . Say the word with the /s/ or /z/ ending:
The word is ______s.
Say _______s.
2. Then make sure you have the student work on pairs.
Say _______[the word without the ending]
Say _________s [the word with the ending]
This part is important because it assures that the student is able to both hear and produce the correct response.
If the ending sound is really a /z/ sound and not an /s/ sound, make sure to model it correctly.
Don’t say runsss. Say runzzz.
Don’t say stopzzz. Say stopsss.
3. Then put the word back in the original sentence and have the student say the sentence three or four times. (or if the student can’t do it after 7-10 trials, stop and pick it up next time) Try to say the sentence in a normal voice. Expect some problems if the word is not at the end of the sentence. The girlzzzare sleeping, for instance, is tricky because you’re really saying one word with no pause between the /z/ sound and /ah/ sound in the word are. (Part of the sentence is actually the word czar.) The same relationship holds for the /s/ sound: The hatsssare red. That’s the way you want the student to say the word in context.
4. A good idea is to use a model word that the student can reliably produce. The reason is that the student won’t generalize what you have taught and you’ll have to correct this particular mistake possibly a hundred times or more. So the simplest correction is to stop the student, say the word correctly. Direct her to say it the way you modeled it. If she has any trouble, direct her to say the model word then go back the trouble word.
For instance if the model word with the /z/ ending is runzzz, and the student has trouble with the word baseballs, the correction would be:
Say runzzzz.
Say baseballzzz.
Got it.
I understand that the part about having to correct it so many times is sort of bitter, but it’s also the truth in most cases.
5. You can use a variation of the steps above to work on any habitual pronunciation problem. With most of them, the model words (and model sentences) are great time savers.
Good luck.
Question 3 comes from Pam Smith:
I’m working with a lady from Micronesia who has great troubles with adding the plural /s/--especially on words such as cats, birds. She also finds it difficult to pronouce /h/ as in hat, horse.
Answer: Part of the answer to your question is in my answer to question 2. The issue you raise about the ending on words like cats and birds is a tough one because it involves very strange sounds when they are said together: dzzz and tsss. Both sounds in dzzz are voiced. Neither sound in tsss is voiced. (If you asked most people whether English has either sound combination, most peopled would say no, but here they are.)
Although it is possible to teach these sounds in isolation (apart from any words), it is probably more efficient to teach them in the context of the words.
Note that these are the same fundamental sound, except for the voice, so you can use the same techniques.
For birds:
Listen: bird.
Say bird.
Repeat the task until the lady is firm.
Listen: birdzzzz.
Make sure you hold the zzz sound for at least a second.
If she has trouble have her say the zzz sound in isolation.
Say zzz.
Then: Say birdzzz.
The big thing you have to watch out for is putting an /uh/ ending on bird. If you’re saying burdah rather than bird, the lady will have a hell of time trying to say birdzzz. She’ll say burdahzzz (sort of 3 syllables).
If she has lots of trouble put a pause between bir and /d/. Pause at least half a second between the sound.
Listen: bir...d.
Say bir...d.
Say bir...dzzz.
Now you’re clearly showing dzzz is a combination of the 2 sounds. Yes, they are distorted somewhat when they are combined, just as tsss is distorted when you say cats.
Say
ca...t
ca...tssss
You can hear how the /t/ sound has changed.
Anyhow, that’s why they have trouble. But understanding why they have trouble suggests how to show them as clearly as possible what they have to learn.
The /h/ sound is easy to teach, but you have to recognize that there are many different /h/ sounds. None of them are voiced: no huh or hah or hih just /h/, like a train puffing. But here’s the problem with the sound. Although it is not voiced, there are a variety of /h/ sounds because they anticipate the vowel that follows them (and in English there is always a vowel that follows them). So practice saying the /h/ sound for the word hit, the /h/ sound for the sound hat, the /h/ sound for the word hoot and the /h/ sound for the word hut.
Then teach it.
When teaching them, model the sound in the word hut three or four times /h/ /h/ /h/ /h/. Make it sound like the little train that could. Then have her say it three or four times. Remember, no voice. When she’s firm, present words that have that sound.
Listen: hut. Say hut.
She’ll do it in a heartbeat.
Do the same thing with the /h/ sound for hot. Then have her say hot. This is something that takes a long time to explain, but a remarkably short period of time to teach.
Let me know how it goes.
Question 4 comes from Almitra Berry:
It is actually not a question, but an expression of a philosophy for teaching English pronunciation. I won’t address most of it because it is general information, but I will respond to parts that have specific instructional implications.
Almitra’s Comment:
If we hammer the ELL about English pronunciation to the point that they lose L1 [native language] proficiency or ability then we are working towards their detriment rather than an additive model of bilingualism.
Response:
Losing this proficiency is about as likely as the student forgetting how to smile or write with a pencil. I know of no data that would suggest such an outcome is remotely possible, and in working with hundreds of students who learn English, I have never seen it. Furthermore, this position is usually expressed by educators or politically-oriented people who declare that we demean the native language when we teach English. In contrast, students usually have a very realistic understanding of their limitations in an English-speaking environment if they aren’t proficient in English. They understand that their ability to speak English has serious implications for employment opportunities and achieving higher SES status.
Almitra’s Comment:
As a rule, if the student consistently articulates an approximation to the English phoneme, we let it go. This is most important if the approximation is consistently tied to oral reading. The most prevalent and readily recognized example of this is Spanish L1 students who cannot pronounce an onset /s/ followed by a hard consonant without adding /e/ in front of the /s/. School is consistently eschool … [other words].
Response:
The least rational starting point for teaching spoken English is oral reading before students are taught to speak in English. In the case of Spanish speakers, the written word prompts them to mispronounce and confuse vowels. In Spanish, the letter E signals an approximation of the long A sound. The letter I signals an approximation of the long E sound. Also, some vowel sounds are absent in Spanish—short a, short i, short u. The letter o signals the long-o sound. Consonants are also a problem: J prompts the sound that is like Y or H in English, and H prompts no sound. Also, students have no sounds for the combination SH and reduce it to CH (and for non-Castilian Spanish, TH becomes Tuh or Duh). Another serious problem is that students transport their Spanish rules about stressing word parts to English. One result is that they leave off endings that English speakers pronounce: /s/ endings, /z/ endings, /d/ endings, /t/ endings. Finally, they don’t have the melody of English. The stress, pitch, and rhythm of English is replaced by a syllable-timed, staccato string of syllables.
The result is that if students learn English pronunciation through oral reading, they usually crystallize their patterns of mispronunciation so it would be virtually impossible to retrain them to speak English reasonably well within less than three years.
Amitra did not ask about efficient practices for correcting the mistake of saying “eschool” for school, but if the goal is to teach English pronunciation, we would certainly not accept the eschool response. Aside from the fact that the mispronunciation is correctable, not correcting the pronunciation has implications about how students conceive of the English article a. A sentence like This is a school actually prompts readers to drop the article, because if they do, the teacher accepts the approximation, eschool. Yet, all they’re doing is using some extension of Spanish rules for not using articles. The English use of articles does not correspond to that of Spanish.
The best approach for teaching students to say the s words is to introduce them very early in the instruction (within the first ten sessions).
Show pictures of action /s/ or objects to make sure students understand what a spoon is, what standing and sleeping are, and what a school is. Then teach pronunciation of these words in isolation. Initially, hold the /s/ sound for at least a second.
Say sssstand.
Say ssspoon.
Say sssleep.
Say ssschool.
It takes some repetition but not as much as it would if you didn’t elongate the s sound. The elongation shows that /s/ is categorically the first sound in these words. When students perform well on the exaggerated words, shorten the /s/ sound.
Say spoon.
Say school.
Say stand.
Say sleep.
When students are firm on this step, present some s words as the first word in sentences. A good routine is to teach commands like:
Stand up. What are you doing? (Standing up.)
Sit down. What are you doing? (Sitting down.)
(Sitting is a good comparison word for standing because it starts with s and follows the same pronunciation rules that standing follows.)
When students are facile, chain these tasks with saying the sentence.
Stand up. [Students stand.]
What are you doing? (Standing up.)
Say: I am standing up. (I am standing up.)
Sit down. [Students sit.]
What re you doing? (Sitting down.)
Say: I am sitting down. (I am sitting down.)
Note that the s words are in the middle of the sentences, which makes the sentences harder than those that begin with the s sound. Note also that the s words are not preceded by the article, a.
Finally, introduce responses that have an article. You use the same routine that you have established for other articles.
What is this? (A ______.)
Say the sentence. (This is a ______.)
When the routine is applied to something like a spoon, students sometimes make the mistakes of saying double articles or no articles.
What is this? (A espoon.)
With what you have taught, it’s easy to correct this mistake.
Say sssspoon. (Ssspoon.)
Say: This is a [pause] ssspoon. (This is a ssspoon.)
Say: This is a spoon. (This is a spoon.)
If the program is set up so that words that start with s + consonant are introduced early, teaching the correct pronunciation for these words is not difficult but requires considerable practice and, at first, students usually won’t generalize to new words.
Almitra’s Comment:
I always ask teachers to be clear on the goals of the instruction. If the goal is for the ELL to read fluently, with prosody, and high levels of comprehension, does it matter if they cannot pronounce with native-like fluency?
Answer:
Possibly it doesn’t matter, but the likelihood of them reading fluently with prosody is slim because prosody is embedded in pronunciation. So students would need considerable practice in saying sentences with all the details in place—the timing, the inflections, and the chains of sounds.
If the goal is to teach students not simply to read but to generate reasonably well-constructed sentences that are comprehensible to speakers of English, the answer is emphatically, Yes, it matters that they cannot pronounce with native-like fluency, simply because that’s a primary goal of teaching spoken English.
Question 5 comes from J. Jackson:
I don’t know if this is an appropriate question because it does not involve non-English speakers. I’m working with a 9-year-old boy who is virtually deaf. He cannot say the sounds of vowels. The mistake he makes is hard to describe. He follows the mouth positions that I show, but then he makes a sound that is something like /n/ or /ng/. I have him hold his hand in front of my mouth, but not much air comes out or he can’t feel it. I have him smile for e and open wide for o, but nothing works. Is there something I can do?
Answer:
Yes. This is not an uncommon response for profoundly deaf children. Demonstrating that the /n/ or /ng/ sounds are not acceptable approximations is pretty easy. Hold your nose closed as you model the /oo/ sound (as in moon). Make sure you hold the sound for at least two seconds. Let him feel the air and your throat. (A lot of air comes out for this sound, and the throat clearly vibrates.)
Then hold his nose closed as he tries to produce the /oo/ sound. It is impossible for him to produce the /n/ sound or the /ng/ sound. So the only way he can match the air coming out and the throat vibrating is to make some approximation of the /oo/ sound.
Once you teach that, you do the same thing for the long /E/ sound, which is harder to shape than /oo/, but doable with practice and patience. Model smiling, holding your nose, and saying the sound. The smiling facilitates producing the /E/ sound because it rules out some other sounds. Holding your nose provides information that it’s a vowel. Note that the student will be able to hold his nose and say a pretty close approximation of the /E/ sound that will have a pronounced nasal component, something like the pronunciation of the /E/ sound produced by a person with a partially corrected cleft pallet.
The correction is to hold your nose closed and have him feel your nose as your say the /E/ sound alternately—with a nasal component and without the nasal component. When you say the sound with a nasal component, your nose vibrates vigorously; when you say the sound without the component, there is no vibration. It’s easy to feel the difference. Use the same technique for the other vowels. In time, he’ll be able to produce fairly good approximations.
(Some profoundly deaf students who learn to speak well have over-learned the rule about no-nasality. This shows up in words that have m or n. They under nasalize these, particularly n, which they pronounce almost as a /d/ sound—no nasality. The word man sounds something like “bad” or how somebody with a serious cold would say “man.” This distortion is correctable by letting the student feel your nose when it is not closed as you produce a proper /n/ sound. The nose vibrates.)
Question 6 comes from Julia Bykova:
What first language to teach a child if the mother and father have different native languages? My prime language is Russian; my husband’s is English. So most of the time I speak Russian with her, but the environment is English, and she will live here. I am a little bit confused because if I teach her Russian, she will not be prepared for school. If I teach her English, she will probably never speak Russian.
Answer:
You are actually in a great position to teach your daughter about language because she doesn’t have to receive a first language, but first languages—English and Russian. I don’t know how old your daughter is, but I assume she’s not yet five and I assume that your husband speaks to her in English and your child responds to him in English.
If this is the case, you shouldn’t have to “teach” English or “teach” Russian but follow a schedule of talking to her that assures she becomes proficient in both languages. Talk as you play, do chores, eat, and other activities.
The reason this strategy will not confuse your child is that the languages have different sounds, different rhythms, different word orders, and different inflections. So she’ll quickly be able to discriminate whether she’s speaking English or Russian. Furthermore, the strategy should not retard her rate of learning. That sounds like an insane assertion, but it’s true. The reason is that probably 99 percent of the time a child uses a language, she is not really learning something new, simply listening to familiar expressions and expressing herself with familiar words and patterns. So if the child is speaking in both languages during the day, instead of 99 percent being familiar, possibly only 97 percent is familiar. But the lower percentage should not make any difference, and the difference will be even closer to 99 if you go out of your way to talk to her more during the day.
Here are the rules:
1. Make sure that she receives a significant amount of exposure to each language every day. Ask a lot of questions. Praise her for smart observations.
2. Insist that she uses the language of the person speaking to her when answering questions or making observations. If you ask her in English, “Where are your shoes?” and she responds in Russian, “V toelete” tell her, “Say it in English”. (In the bathroom.)
3. When you’re doing things with her, switch from one language to the other from time to time. Tell her, “Let’s talk in English for a while.” The value of this procedure is that it pretty much assures that the child will learn all the vocabulary for both English and Russian associated with the same activities.
4. Don’t get overzealous about correcting her speech. The reason is that if you correct her too much, you’re actually punishing her for talking, and she’ll tend to cut down on how much she talks. So don’t make activities “teaching drills.” Make them experiences that interest her, so she wants to talk about them.
5. If she encounters something she doesn’t know how to say, tell her what it means, have her say the words a couple of times, and make sure that she uses them in your conversations. Later, ask her if she knows how to say them in Russian. For instance, she learns the meaning of “some things are dangerous.” You discuss things that are dangerous. Later you ask, “Do you know how to say “Some things are dangerous” in Russian? (Neskol’ki veshchi opasny.)
6. Praise her for asking questions about how to say things in one of the languages. When she asks these questions, she is showing you that she is comfortable with the idea that she speaks two languages and wants to keep her understanding balanced.
Follow these rules and she’ll do fine in both language.
Note: I can’t speak Russian, but my son Kurt can. He told me how to write the Russian words.
Q1-/r/ sound | Q2-final /s/ | Q3-/ts/ /h/ sound | |Q4-Spanish Speakers | Q5-Deaf Students | Q6-Primary Language |
Submit your questions to zig@nifdi.org