The Goldlist Method and Pronunciation of a Learned Language.

The chinese word for "National Language&q...
The Chinese word for "National Language" in Traditional and Simplified (ugh) Characters, Pinyin and 3 outdated transcription methods.

I haven’t always been as explicit as I could be on the matter of pronunciation of languages, and how the Goldlist method works for that. A great question from Mitch which I saw today puts his finger on what learners could perceive to be a difficulty while using the Goldlist method and how best to approach this question. 

First his question, which you can see in the page on the Goldlist Method.


1 question. The gold list is activated after 3 days in an immersion environment. This will require correct pronunciation. How do you go about getting this for each word? Do you find audio for each word on your list? I don’t remember doing this as a child. I don’t think I asked someone how to say this for every word I found. I’m missing something.

Thanks Viktor.


This is actually a very good question, so in addition to answering it here I think I’ll also make this answer a fully fledged article on the front page of the blog. I’ll kick it off here and continue on the full article.

The first thing I will take issue with is your statement “this will require correct pronunciation”. I am not sure what “correct pronunciation” is, all I know is that there are people who mimic native pronunciation better than others and they may sound like better linguists when what they probably are is just better voice actors.

What you definitely need to have an awareness of is how that word is supposed to sound so that you would be able to say it understandably – without having a native listener confusing what you were trying to say, and to recognise the word when another person says it.

You don’t need to worry about this mythical holy grail of “correct pronunciation” in the way you’ve formulated it in the question.

Now for most languages in the world, and surprisingly not for the so-called “easy” language of English, one of two things is true. Either the normal way of writing is a lot more phonetic and devoid of exceptions than the English is, or in the cases where this is not the case there are ways to write it more phonetically.

Let’s give some concrete examples.

Let’s take German as a good example of a language where, once you’ve learned a handful of rules as to how the pronunication of the letters differs from English, and where the stress is likely to fall, the difference in “ch” when it follows i or e from the way it sounds after a,o or u, how the umlauts sound and the “beta s”, then any word you see in German, you’ll be able to pronounce accurately. For the sake of argument let’s call that a Group A language.

For Group A languages, you only need to learn the rules and you can accurately predict both the pronunciation of words you see, and with more practice you can also make a good fist of writing down the word you hear and spelling it right. Besides German you could put Italian and Serbian, Greek and at a pinch Spanish and Czech.

Now let’s call Group B languages such languages where you can predict the pronunciation pretty well from learning a finite set of rules about how the letters sound in various combinations, but you would have difficulties sometimes in working out which of two or more ways to write a sound you hear. French is a good example, and so is Polish and I’d say Danish because of its tendency to swallow sounds. Whatever you read in these languages you should be able to make a good job of pronouncing once you’ve learned a finite set of rules (finite enough to be covered within your first 1% of study) but it will take a while to predict exactly how a new word you hear will be written – and native speakers also will have that problem, usually only while they are kids.

Now let’s consider another grop of languages and call them Group C. These are the languages where you get the basic idea from the way its written, but some information may be not available, you have to know it. In Russian this would be the stresses and the ye/yo. Which is why learners’ books have the stress added and always have the two dots on the “yo”. If you consider Thai, it is probably also Group C even though it has different consonants for different tones, because you are fitting five tones into three consonant series. You get a load of help from that, but it doesn’t get you the whole way, you still have to know a lot.  A stronger Group C scenario is where you’ve got missing vowels, as in Arabic or Hebrew. Again for learners the textbooks will have the vowels added in additional dots and dashes above or below the letters, but you won’t see these in advanced texts or in texts written for people who already know the language.

English deserves to be considered as a tougher contender than your usual Group C language, because whole swathes of vocabulary have unpredictable pronunciations from the spelling, and also there are numerous homophones, and we never write English even in learners’ books the simplified way – (duing so wud bee considered simply rong) so that creates more issues than for learners of Semitic languages.  In some ways, however, English is more straightforward and for many words it’s like a Group B language, so I’ll leave it in group C for a balanced view.

Group D would be the languages where you don’t get much of a clue at all from the writing as to how things sound. Archetypally we are talking about Japanese and Chinese. There are not really any further languages that function on a full scheme of hieroglyphics left. Korean can use Chinese hanzas and many Koreans learn them, but what you are seeing when you look at Hangeul is an alphabet system, but one which connects in two directions, not just one, giving rise to beautiful decorative “characters” – which are not really characters at all. Similar to this is Tibetan which is stacked, and a number of other Asian alphabets could be classified in that way, but these are not group D really – some of them are actually Group A by this classification, just unfamiliar looking/sounding Group A.

But even Group D languages offer us tools while we are learning. Japanese supports the learner of kanji with kana and you can also find plenty of Roomaji support (if you have trouble with the Roomaji, look out for Professor Huliganov’s short book on this blog somewhere “Remembering the Romaji” and pinyin is there to support the learner of Mandarin. And Japanese written in Romaji and Mandarin in pinyin are back in Group A, because you can both accurately pronounce a Mandarin or a Japanese word from these respective methods as well as accurately transcribe a word which a speaker speaks clearly and properly into those systems.

Now having offered a classification of languages as A, B, C and D, we can make subgroups 1 and 2 for each of these depending on whether there are lot of homophones in the spoken language or not so many. Not so many would be subgroup 1 and many would be subgroup 2 in each case. There is no such group that I’m aware of as D1, as Chinese and Japanese are both full of homophones and so they are D2 (or A2 when written just in their kana/pinyin forms).

A further way to classify pronunciation difficulty is how different the sounds are to the ones you are already capable of making and distinguishing, but this depends on what language you are coming at the new language from. What is hard about pronouncing Chinese for an American is not the same as what is hard about it for a Pole. It is rather easier for a Pole, they have all the sounds of Mandarin except the “r” – but here the American has it easy as Mandarin has “r” similar to the typical American pronunciation of that letter. So I can’t make a global classification on that basis. Also some people can learn new letters with ease and can get Xhosa clicks the first time they try, whereas others can’t manage to roll their “r”. So I’m not going to even try and classify this part of pronunciation difficulty.

Let me now apply this classification to the 30 top languages I predicted would be of highest economic value to the learner in 2050.

    A B C D
L     F
2 English     F  
3 Arabic L   F  
4 Russian L   F  
5 Spanish F      
6 Japanese L     F
7 German F      
8 French   F    
9 Portuguese   F    
10 Korean F      
11 Italian F      
12 Turkish F      
13 Hindi   F    
14 Dutch   F    
15 Malay/Indonesian F      
16 Polish   F    
17 Scandinavian languages   F    
18 ex Serbo-Croat group F      
19 Hebrew L   F  
20 Thai with Lao     F  
21 Czech with Slovak F      
22 Swahili F      
23 Vietnamese F      
24 Ukrainian     F  
25 Greek F      
26 Farsi   F    
27 Romanian   F    
28 Tagalog F      
29 Bulgarian F      
30 Hungarian F      
  Totals – full literacy 14 8 6 2
  Totals – books for learners 19 8 3 0

So the conclusion here is that almost all the languages people are likely to want to learn most are either easier than English with regard to predictability of pronunciation or else they are available with learners’ tool like the accented Russian texts, the fully vowel pointed Hebrew and Arabic texts and the Chinese character learning books which come with a variety of fully phonetic solutions.

English is really less phonetic than the norm. Which only means it has a clearer open door to its rich linguistic history than it would have had if we had simplified it.

So really the issue of being able to predict pronunciation from a written word is a lot easier in most cases than you possibly think it will me. Certainly the rules for learning the group A and group B ones will be finite. How quickly you learn them and how much of the whole learning task they make up ( less than 1% to say 3% in extremis) depends on your ability with different alphabets and in making unfamiliar sounds. But the rules, in so far as you can write them as written rules, can all be goldlisted and learned by the gold list method.

Even if the language looks easy, you may want to give yourself the best possible accent and not sound more foreign than you have to. The best way to achieve this is to put some time into an audio only course before you even start to look at how these languages are written down.

That means that before even beginning a gold list method you could invest in a Michel Thomas and/or a Pimsleur course. These are purely auditory courses. You can do them entirely walking around with earphones on. The correct time to do these courses is right at the start before you do a gold list.

Michel Thomas method courses are really better than Pimsleur, but aren’t available yet in all languages, there are about a dozen, whereas Pimsleur’s pretty much got the list above covered.

A Pimsleur course consists of typically 30 half hour lessons, and they don’t really explain structure, they give you an almost phrase book approach. They seem typically concerned with American men going to different countries trying to get dates with local women, which can become tiresome after a while, where as all the MT method courses are unique.

A Pimsleur gives you 15 hours material, and a full, three stage Michael Thomas course gives you 16 hours of material roughly, over three phases; foundation, “advanced” and “vocabulary builder”. The total material that they give doesn’t really qualify as “advanced” but it is an advance on the “foundation” so in that sense it is. But if you have a language where there is both a Pimsleur and a Michel Thomas course, the best approach will be to kick off with the whole Michel Thomas course, and then do the whole Pimsleur one, which will give you a great sense of the sounds and structure of the language before you even see it written down.

Working through 30 hours material at an average of say an hour a day will mean that the first month will go by without you starting the gold list. But the amount in these audio courses is very finite, and the goldlist method is able to be used for much more volume as you start ot off on a beginners’ written course book, go through intermediate and final course books so that you have maybe 2000 – 2500 words just from courses, whereas the Pimsleur and the MT combined will be unlikely to get you beyond a thousand and you won’t know how to write them. But what they will mean is that when you see them written you’ll know how to pronounce them properly, and the experience will be more similar to, as you mention, the way you learned as a child. You didn’t need to ask for the audio on most words, you learned them actually as audio subconsciously, and may consciously afterwards have learned the spelling, not so much the other way round, although almost certainly that will have happened on some words. This is done for children by making them read aloud.

In summary – the correct place to fit an audio only course around the Goldlist is to do the audio only first, and the goldlist afterwards. And it’s definitely worth doing.

Don’t buy the abbreviated “taster” courses of MT, though – it’s material which is the first two lessons of the foundation course anyway. Assume you will take the whole course and save your money on that. It’s not like they’re cheap. At least the goldlist method later means you can save on expensive and pointless teacher – based lessons for most languages.

Many thanks for that question, Mitch. Those others of you who recently asked questions on this blog, don’t worry, I’m getting to you as I can.

4 thoughts on “The Goldlist Method and Pronunciation of a Learned Language.

  1. Thank you for the great explanation. Your right about the “mythical holy grail” of correct pronunciation. I completely agree. I’m a Southerner in the United States so I know all about butcherin’ English pronunciation.

    I guess, I meant to say as close as possible to correct pronunciation to be understood.

    You’re much appreciated. Thanks

  2. Wow, thanks a lot, this article is very clarifying.
    I was curious about this too, but I didn’t ask.

  3. This is a very fascinating read. Your overviews of languages in general open up new worlds for me.
    A great pronunciation help for me has been Not only does their Russian dictionary contain more words than my Oxford, but by clicking on the word I can hear the proper pronunciation. It’s always delightful when I compare my pronunciation with theirs and find it to be correct, and it causes me to investigate further when I’m wrong. Of course watching online Russian television helps,but doesn’t expand the vocabulary much.

Your thoughts welcome, by all mean reply also to other community members!