Steve K or Mxsmanic?

Tom   Wednesday, October 13, 2004, 00:45 GMT
[Steve, Mxsmanic -- Please forgive my simplistic descriptions. I hope you can clarify your views below.]

Two smart and successful language learning theorists -- Steve K and Mxsmanic -- came to the Antimoon Forum about a month ago. Their views are quite similar to those presented on Antimoon. However, there is one area where they differ quite conspicuously.

Steve K is what I would call a "naturalist". He believes that learning a second language should be as similar as possible to learning your first language as a child. Whenever you use technical grammatical or phonetic terms, Steve K gets nervous. "What a waste of time!" he yells. In his view, the ideal language learner should not know what "present perfect" means or know the phonetic alphabet, for these are, in his own words, "artificial barriers between the learner and language".

In Steve K's opinion, spending 15 minutes to learn about the three types of conditional patterns used in English would be a waste of time. You would spend the time better if you read a couple pages with English sentences, he says.

By contrast, Mxsmanic is more of an engineer -- an English teacher who often brings up grammar-related points in his classes. He points out that children take a long time to fully master a language (up to 10 years) and that adults do not have that kind of time. Therefore, he advises learners to take a shortcut and produce sentences by following grammatical rules. He believes every English sentence is perfectly logical and can be explained by such rules.

He is also a big proponent of phonetic transcription. He claims to have developed his French pronunciation to a near-native level mostly by studying phonetic transcriptions. He even says that you can learn the pronunciation of a foreign language JUST by studying transcriptions and without listening to one recording.

Mxsmanic gets quite technical at times -- for example, ordinary phonemic transcription (such as you might find in a dictionary) is not good enough for him. He prefers narrow phonetic transcriptions which show precisely how the phonemes are realized in connected speech.

Now where does Antimoon stand? As I've mentioned, both gentlemen's opinions are very similar to the views presented on Antimoon. In regard to the differences described above, I believe I stand closer to Mxsmanic than to Steve K.

1. I believe it is useful to learn some grammar terms. When you can name something, you have control over it. You also notice it easily. Learning the fundamentals of grammar allows you to learn more from every sentence you read, because it makes you pay attention to things which are potentially problematic. If you know that some verbs take the gerund and some take the infinitive, you pay attention to such verbs in the sentences you read. If you don't know it's an issue in English, you don't pay attention to it, and you don't learn to do it properly.

2. Similarly, I believe it is useful to learn the basics of phonetics, for the reasons I presented at
I can't imagine learning the pronunciation of a foreign language without learning what sounds it uses. Even if all you do is listen to recordings, you have to realize that "cat" has a different vowel from "cut". You can call it "the vowel in 'cat'", but why not learn a handy, widely understood symbol (@)? Learning the symbols is a one-time effort that will allow you to use dictionaries more effectively, make pronunciation-related notes, etc.

[BTW: I wonder how Steve K refers to the different vowels used in German?]

At the same time, I am not in complete agreement with Mxsmanic. I certainly never used phonetic transcriptions of connected speech when learning English pronunciation. I found phonemic transcriptions of words and imitation of recordings to be quite sufficient.

I do not believe that grammar rules are anything more than an unwieldy, inaccurate crutch for learners who have not internalized enough input. I occasionally use grammar rules when writing, but I invariably find that I can't use the same rules when speaking -- using them is just too time-consuming. Besides, they are hard to memorize and always have exceptions. See also

I do not believe that grammar rules can explain everything. Sure, they tell you how to express your thoughts in words, but they do not tell you how you're supposed to think in order to speak in a natural way. And the way native speakers think can be substantially different from the way foreigners think. For example, as a Pole, I think of "police" as singular, while native speakers of English think of "police" as plural. A matter of convention, not logic.
Joe   Wednesday, October 13, 2004, 01:16 GMT
I look at it this way. How many people in the world can speak their native language but aren't literate? Phonetics and grammar rules are especially important if you plan to master a language, meaning be excellent in composition and conversation, as well as be grammatically correct in both forms of communication.

I can't tell you at the number of American adults who cannot, even after years of schooling, compose a proper essay or commit grammatical fallacies.

It all depends on what you're looking for. To me, I feel that the base is to learn how to speak the language, and then it can be supplemented with grammar and phonetics. In fact, I NEED grammar rules and phonetics to help me make rhyme or reason of some things. Learning German I looked up German grammar rules many times just so I have a comprehensive understanding of why you say things the way you do.

Learning a language as an older child or adult is different than learning your first language, because at this point in life you understand the sentence structures and grammar of your own language and so you can bypass some of the rudimentary trial and error of your first language experience. However, much of the basics prove true. You do learn best through speaking, and I personally don't understand how anyone could learn to speak perfectly without ever once hearing the language's sounds.
mjd   Wednesday, October 13, 2004, 02:12 GMT
I agree with Mxsmanic's approach with respect to formulation of sentences and learning tenses. However, I think the best way to learn pronunciation is by imitating native speakers...not just scratching the surface of pronunciation, but by making a conscious effort to mimic the accent as if one were doing an impression of someone.
Mxsmanic   Wednesday, October 13, 2004, 03:16 GMT
The advantage to a "naturalistic" method of language learning is that it eventually produces native proficiency, free of any errors. The disadvantage is that it takes many, many years to do this. Children learn languages in this way and eventually achieve perfect native fluency; but they require two decades to do it, practicing all day long, and during that interim period they make lots and lots of mistakes.

The advantage to a more formalized method of language learning, including the study of grammar and phonology, is that it provides a shortcut to a high level of proficiency in a short time. Native-level proficiency is no more quickly achieved than it would be with naturalistic learning, but a student can get closer to that level, faster, by deliberately studying rules in order to skip the trial and error process that requires so much time. The disadvantage to this method is that the proficiency attained is less permanent and complete than that attained through laborious trial and error in a naturalistic setting. People who learn languages formally can come very close to native proficiency in only a few years, but they tend to retain certain mistakes indefinitely thereafter. For many people, this is acceptable, as the goal is to communicate, and not to pass for a native speaker. Those who wish to achieve native fluency must eventually resort to more naturalistic methods to remove those last few mistakes and errors, however, and this requires some years of practice (ideally with total immersion in the language, if possible).

For ESL students, the formal approach tends to be more in line with their personal goals. They wish to communicate effectively in English, and they wish to do so as quickly as possible. The shortcuts provided by deliberate study of rules allow them to reach a functional level very quickly indeed, even if these shortcuts are not quite as useful for attaining native proficiency (which still requires a certain amount of practice). ESL students don't mind speaking with an accent as long as they are completely fluent for their purposes. Thus, if I teach ESL students, I favor this formalized approach.

Those who wish to achieve perfect native fluency ideally would learn a language through decades of total immersion, as children do. Nobody has that kind of time in real life, though, so formal study can shorten the learning period, and then a naturalistic approach can bring the student towards the perfect proficiency that he desires once the basics are in place. With any formal study, there's always the risk of "backsliding," because one has learned rules and has not ingrained the proper use of the language through tedious trial and error; but if formal study is used to achieve high proficiency at an early date, subsequent years of practice and naturalistic study and exposure should "fill in the gaps" and eventually produce the same native proficiency that purely naturalistic methods would have produced, ideally saving some time along the way.

One other consideration is the influence of one's first language. Some say that the greatest obstacle to learning a second language is having a first language. ESL learners are comfortable in their native languages, and this is a huge temptation to avoid study of English in depth. One of the greatest challenges for ESL learners is to apply themselves to English with the same dedication that they applied as children to the learning of their native languages. If they manage that, they'll eventually speak English perfectly--but very few people manage it, because they lack the motivation, the discipline, or both. And in many cases it just isn't necessary to achieve the student's goals, which are usually less ambitious. Both the naturalistic and the formal methods of study are handicapped by this reality. For learners of a language who already know another language, motivation is the number one determinant of success, with discipline, open-mindedness, and intelligence following close behind (not necessarily in that order).
Beginner   Wednesday, October 13, 2004, 04:07 GMT
I like Steve K's approach.
|||   Wednesday, October 13, 2004, 13:42 GMT
k WEb@!
Steve K   Wednesday, October 13, 2004, 14:15 GMT
I have been away for a week or so and what do I find on my return.

I think that antimoon, Mxsmanic and I can agree on most areas of language learning. Here is my take.

1)Use authentic content and let the learner choose what interests him or her. Keep the learner motivated. Make it about genuine communication. Give them lots of input and do not worry about performance. Performance will come once the learner had develop a feel for the language.

We should develop massive corpora of authentic language for learners to choose from in all languages. All the language textbooks and the time teachers spend finding or creating content is unnecessary. The trick is just to find a way for the learner to evalutate the difficulty of a given content item and to find what they want.(I admit some learners prefer to be given a course of study and that is easily arranged).

2) Always combine audio and text.

3) Focus on vocabulary (words and phrases) learned in context, over grammar.

4) Now here is the kicker. The learner must read. The learner must observe while reading. The learner must save words and especially PHRASES. In my experience the learner gradually develops the ability to see phrases that he or she wants. As they get better at using phrases rather than translating from their native language, their new language becomes more natural. And we feel we have a system that greatly helps in the learning of these words and phrases saved from the text that the learner is reading and listening to.

5) Write. That is the quality control mechanism. There the teacher provides some grammatical explanation, but mostly replaces poor phrases with good ones.

6) And pronunciation, ..just imitate. IPA is fine when you do not have the kind of sound file management systems, MP3 players, CD walkmen etc that we have now. Pronunciation in any language can vary. Just imitate the accent you want. Work on it often, in words and phrases and sentences.

I look forward to the views of others.
Mxsmanic   Wednesday, October 13, 2004, 14:35 GMT
Imitation is an extremely slow way to acquire correct pronunciation for learners who already speak a first language. We grow accustomed to the phonemic differences in our native languages and we learn to listen for them and hear them; but at the same time, we learn to ignore and not hear differences that are not phonemic in our own languages, even when they are phonemic in other languages we are trying to learn. Worse yet, we hear differences that are irrelevant in other languages if they are phonemic in our own, and this confuses us in the learning of other languages.

Formal study of phonology can remedy this, and greatly reduce the time required to acquire good pronunciation. By "greatly reduce," I mean a ratio of perhaps ten to one. Quite a bit of bang for the buck.
Tom   Wednesday, October 13, 2004, 22:24 GMT

"The disadvantage [of the naturalist method] is that it takes many, many years to do this. Children learn languages in this way and eventually achieve perfect native fluency; but they require two decades to do it"

First of all, the fact that a child needs 8-10 years (20 years seems excessive) to learn a language well does not mean that an adult needs the same amount of time. For one thing, I am much more intelligent that I used to be when I was 1 year old.

I took my English from poor to impressive in 2-3 years with hardly any grammar study.
- I was exposed to some grammar rules in classes, but I never managed to keep them in my memory.
- I occasionally formulated my own grammar hypotheses when reading, but they were secondary to my "grammar intuition". I was more like a native speaker who feels a certain way about certain structures and tries to come up with a rule for how he feels. For example, I would feel that the present perfect is appropriate in a sentence like "I have been to England before", so I'd think "I guess that's because I'm really talking about the present (my current "database of experiences") and the moment in time when I went to England is unimportant".

As I described at (scroll to end of article), when I first opened a grammar reference, I was positively surprised that there were so many formal rules for things I found obvious due to all the reading I had done.

- I did learn the IPA and use transcriptions in dictionaries extensively. I did not study phonetics beyond that.

Of course, I already knew some English when I started the 2-3 year period of intensive learning, but I still say it's possible to learn English in 3 years with the natural method.

"in order to skip the trial and error process that requires so much time"

What do you mean by "trial and error"? I don't remember using the trial-and-error method when learning English. I just imitated correct phrases. Sure, I made some mistakes, but it's not like those who build their sentences by following grammar rules make no mistakes.
Tom   Wednesday, October 13, 2004, 22:33 GMT
"We should develop massive corpora of authentic language for learners to choose from in all languages."

Steve, those corpora already exist. They're called books, TV, movies, and the Internet.

"Always combine audio and text. "


"Focus on vocabulary (words and phrases) learned in context, over grammar."


"The learner must read. The learner must observe while reading. The learner must save words and especially PHRASES."


"5) Write. That is the quality control mechanism. There the teacher provides some grammatical explanation, but mostly replaces poor
phrases with good ones."


"And pronunciation, ..just imitate. IPA is fine when you do not have the kind of sound file management systems ..."

I have a question: You said you speak German well. Are you aware of the sounds used in German? If so, how do you refer to those sounds?
Tom   Wednesday, October 13, 2004, 22:52 GMT
"we learn to ignore and not hear differences that are not phonemic in our own languages"

Exactly. Knowledge of phonetic transcription sharpens your ability to learn pronunciation from recordings. You begin to notice differences between sounds that you thought to be identical.
Steve K   Thursday, October 14, 2004, 04:59 GMT
Audio alone is not good enough because

1) you need to understand what you are listening to
2) visual and aural learning are mutually reinforcing

Text alone is acceptable for advanced learners but earlier learners need to listen a lot, a lot, and need to understand what they are listening to, i.e. need the text. So combining both is very effective.

TV internet books etc. this is fine, as long as it is accessible in audio and e-text format. What needs to be added is natural interviews and discussions with real people, with transcripts. Yes it would not be difficult to make this corpus available to learners as the only text book they need. All that is needed is the system that enables the learner to use it properly. ( I.e. our system)


I never bothered with any phonetic script, nor did I for any other language. Just listening in a systematic way, with an open mind to individual sounds and sentences, and then to longer more interesting texts, over and over again.

With English , given the irregular spelling, it is necessary to collect words with the same vowel in the dominant syllable and listen to them together. I do not see the need for 10 years etc. but I do not deny that the IPA may help some people, I just have not seen that a familiarithy with IPA correlates to good pronunciation in people that I have seen.
Steve K   Thursday, October 14, 2004, 14:46 GMT
RE IPA. Yes it can be useful at first, or for new words, for those people who are familiar with it. However, the problem that most people seem to have with pronunciation is either intonation or the inability to make certain sounds."Th", or "w" or the aspirated "h", or the difference between r and l, or the "77 sea sick sailors" in Swedish, or the final consonant for Chinese speakers etc.So the IPA cannot help there.

On the other hand a systematic exposure to lots of listening, plus deliberate listening and recording of words and sentences and listening to oneself, if practiced daily works well. We took all the most common 650 words in the language and grouped them into strings of 5 words with the same vowel sound in the dominant syllable (regardless of spelling), or with key consonant and double consonant sounds, for repetitive listening and recording by the learner.
Mxsmanic   Thursday, October 14, 2004, 19:40 GMT
Trial and error is the method used by children learning their first languageā€”because they have no other option. I don't think anyone uses trial and error after acquiring their first language, as just about anything is more efficient than that. However, trial and error does provide perfect fluency, eventually, since by its very nature it forces the learner to experience and correct every single error (hence its name).

Most children do not speak their native languages perfectly even after a decade, unless they've been given formal instruction in those languages. And many people are still making mistakes after two decades.

BTW, intelligence and aptitude are pretty much fixed throughout life. One can acquire knowledge, but intelligence neither increases nor decreases significantly. Someone born stupid will always be stupid, and someone who is smart in adulthood was just as smart as a toddler.
Tom   Friday, October 15, 2004, 11:24 GMT
"Text alone is acceptable for advanced learners but earlier learners need to listen a lot, a lot, and need to understand what they are listening to, i.e. need the text."

I recommend that beginners focus on reading instead of listening, because reading is easier. Once they acquire some vocabulary, they can start listening.
I don't deny the effectiveness of the audio/text combination, but most fun content is either text or audio (not both). There are two notable exceptions:
- Movies on DVD (with subtitles)
- Adventure games (titles + voice)

On the whole, I'd rather have audio-only or text-only content that is fun than be stuck with a limited library of boring content that's designed for learning English.

"Yes it would not be difficult to make this corpus available to learners as the only text book they need."

The Internet is already available as are books and movies.

About German, I wonder if you can classify the following words into two groups, based on the vowel they contain?
[This question is for Steve K only]