How to learn English

I am an English teacher (primarily of English literature) who currently runs a tutoring company in Hong Kong. However, I have also taught students to improve their English, strengthening students who are intermediate or even beginner learners of English to become much stronger.

Your Antimoon resource is quite valuable. What you teach there accords with my findings. The biggest thing students need is good English inputs, and those indeed come from reading, especially.

Something that you may want to consider drawing more explicit attention to: it’s worth encouraging students of English not only to read, but to read aloud.

The problem with just reading quietly is that it does not necessarily help you improve your grammar and expression that much. As an example of this, I had a professor at university whose third language was English. Although he could publish his articles in academic journals, he still made some grammar mistakes when speaking, and undoubtedly when writing, too.

By contrast, when you read aloud, you are actively engaging with the material. You are practicing physically producing the different words with a proper syntax and correct expression. Over time, the gradual result is that both your spoken and written grammar will be naturally improved, because you are both practicing correct English grammar, and hearing yourself as you speak it.

Just as an anecdote to illustrate this, I am currently helping a student who goes to a local school in Hong Kong (where the style of English teaching is pretty bad – lots of grammar-focused instruction, rather than genuine reading). She also rarely needs to use English at home, further hindering opportunities to practice. Nevertheless, our lessons have focused upon lots of reading aloud. She is becoming better at naturally correcting sentences to include things like articles, which, as you may know, simply don’t exist in Chinese, whether Cantonese or Putonghua.

I think it would be worth adding this recommendation to your website. I know lots of students struggle with grammar and good English expression. Reading aloud is a way to fairly rapidly improve, depending upon how frequently one reads aloud. If someone read aloud for 15 minutes a day for about a year and a half, the difference in their grammar and expression would be dramatic.

When I was learning English, I didn’t use reading aloud as a technique to improve my grammar, so I cannot attest to its effectiveness. For me, what worked best was being mindful of the grammar in the English sentences that I read or produced. Simply taking a moment to notice that someone said he went to the doctor, rather than a doctor, for example.

Speaking purely theoretically, I am not convinced that reading aloud is a road to grammatical progress. In my experience, reading aloud can put your mind in a kind of thoughtless state where you simply pronounce the words like a robot, without thinking about the meaning, never mind syntactical nuances. When I read aloud, I certainly feel that I can understand less of what’s on the page, probably because my brain is busy recalling how to pronounce each word. Am I the only one who’s had this experience?

It is perhaps worth mentioning that reading someone else’s words aloud is a very different activity than speaking. The latter involves doing the actual hard work of choosing the correct grammatical structures and putting them together. If you can avoid making too many mistakes, speaking will let you learn much more.

Is there something magical about physically producing sentences, something that outweighs the “mindless robot” effect that I discussed earlier? There could be. Perhaps it would be worth combining quiet, mindful reading with reading aloud. Or perhaps quiet, mindful reading is always better.

• Where do grammar rules come from? How do they differ from the rules of formal systems like programming languages?
• Can grammar rules describe a language completely?
• Are there any good ways to use grammar rules?

Why you shouldn’t rely on grammar rules

Why is it so difficult to learn a language? The main reason is that speaking a language correctly requires a vast amount of knowledge – far greater than is necessary to be a competent doctor or lawyer. A large chunk of this knowledge is, of course, vocabulary. To speak English fluently, you have to know the meanings and pronunciations of at least 10,000 words and phrases (for comparison, the average college student in the US knows about 20,000 words).

But while most learners realize vocabulary is a major area that requires a lot of attention, fewer are aware that there is an equally large body of facts that is described with the word usage.

Read more

It is natural to assume that phonetic transcriptions are precise instructions on how to pronounce English words. I remember a time when I thought so, too. Over the years, I have realized that they are more like statistics: they can lie, and when they don’t lie, they don’t tell the whole truth.

In my new article (complete with audio recordings), I explain why phonetic transcriptions don’t give you the whole picture and can even be harmful if you take them too literally.

Read more in: “Things you should know about phonetic transcription”

Many languages have sensible writing systems. If you look at a Spanish, German or Italian word, you can tell how to pronounce it – all you need to know is a handful of basic rules. But English is not one of those languages. English words with almost identical spellings often have different pronunciations, so looking at a word’s spelling doesn’t tell you very much.

In my new article, I try to pinpoint why English pronunciation is so difficult that even native speakers occasionally get in trouble. I put on a horror show of tricky pronunciation examples that will make you abandon all hope of learning to speak English without mistakes. Finally, I offer a ray of hope as I explain what you can do to survive.

Read more in “English pronunciation is a minefield — here’s how to survive”

The article includes audio examples and a brief explanation of why certain sounds seem different while others seem similar.

Read more in “Why it’s important to learn about the sounds of English”

• Click / tap / point cursor at any highlighted words to get a pop-up explanation (with additional links to online dictionaries). (Try it right now!)
• Select any word or phrase to get a popup with links to the best online dictionaries, enabling you to quickly look it up. (Try selecting something.)
• Completely new audio links and audio player code built around HTML5 and SoundManager — does not require Flash, works great on mobile devices. (Examples of new audio links and new audio player)
• Wider layout for more comfortable reading of articles. More space for sidenotes (of which there are many) and large images.
• Responsive layout for mobile devices (including blog, forum and wiki pages). Antimoon now works pretty damn well on tablets – in fact, probably the best way to read this site is sitting in an armchair, holding a tablet. As for smaller devices like smartphones… well, it works as well as it can. You’re not going to get great reading experience on a 5-inch screen anyway.
• Much faster performance on blog pages and translation wiki, thanks to: optimized templates with fewer HTTP requests, optimized image size, database tuning, PHP cache & updated software.
• A bunch of audio files have been re-recorded on new equipment for better quality.

New articles coming soon.

1. Pronunciation dictionary. This works for 90% of words.
2. Forvo. This has the advantage of giving you authentic pronunciations free from possible biases.

I was recently made aware of a third option – a website called YouGlish.com, designed by Dan Barhen. How does it work? There’s a huge mass of content available on YouTube. Some YouTube videos include transcripts. YouGlish lets you search those transcripts. Type a word, press Enter, and you get a video where someone is saying that word. The video will automatically start from the relevant sentence, so there’s no need to look for the right place. If you want more examples of your word being pronounced, simply click the big arrow button to go the next video result. You can also easily replay your sentence. It’s a very nice interface for browsing examples or real-life speech and congratulations are due to Dan for making it work so smoothly.

Forvo and YouGlish are great tools for the linguistically inclined who don’t trust their dictionaries. Dictionaries can be biased, outdated, or mistaken; no dictionary will beat direct access to real-world usage, which is the “source data” dictionaries are (or at least should be) based on.

Unlike dictionaries, these sites often enable you to hear the same word pronounced by dozens of speakers, which allows you to get rough answers to statistical questions like “What percentage of Americans pronounce a syllabic L in words like idly and peddling (making them sound like idally, pedaling)?”, “What percentage of Americans pronounce quarter as though it was spelled corter?”, etc.

Anyway, if you have Spanish-speaking amigos who keep asking you “¿Cómo aprender inglés?“, do them a favor and point them to the Spanish wiki.

(…) There are different types of transcriptions for American English (and for British English). All of this is confusing because I am not absolutely sure whether they describe one and the same pronunciation or they describe different pronunciations.
For instance, the word “air” in American dictionaries is transcribed in many ways:
1. /ɛər/ (this transcription is clear: it means that there is a schwa sound and the “r” sound is pronounced)
2. /ɛr/ (this transcription is not very clear to me: is the pronunciation the same as the transcription above or does it mean that the schwa sound is omitted/deleted/?)
I have noticed that a lot of American dictionaries use the /ɛr/ transcription. Do they mean that the schwa sound is omitted or does it mean that /ɛr/ is the same as /ɛər/?
Also, in different dictionaries the word “near” is transcribed as /nɪər/, /nɪr/ and /ni:r/). Do they mean the same thing?
And in different dictionaries the word “sure” is transcribed as /ʃʊər/, /ʃʊr/ and /ʃu:r/. Do they mean the same thing?

Yes. Whether air is transcribed /er/, /ɛr/, /eər/, /ɛər/, /eɚ/ or /ɛɚ/, it refers to the “standard way air is pronounced in General American English”. Note that the “standard way air is pronounced in General American English” is a slightly fuzzy concept – a number of slightly different pronunciations can be called “standard”.

I don’t think you should assume that dictionary editors prescribe any particular way to pronounce air. Dictionaries have to choose some convention – otherwise, they would have to transcribe air as /er | ɛr | eər | ɛər/. Remember that the goal of phonetic (or, more precisely, phonemic) transcriptions in dictionaries is to help you group words into classes, so that you know that where belongs to the same class as air, but were does not (even though it has a similar spelling). Transcriptions are not detailed instructions for producing the actual sounds. (As I wrote in my latest update, you shouldn’t take them too literally – instead, you should focus on listening.)

Another thing worth realizing is that people who design phonetic transcription systems for dictionaries like to use the smallest possible number of symbols to keep things simple. So they will transcribe both bed and bear with /e/, even though bear is usually pronounced with a raised vowel (“raised” means “pronounced with a hint of [i]”). They just count on you to know that if /e/ is followed by /r/, you should raise it a bit. They can get away with this because this “raised e” vowel always occurs before /r/ and never before other sounds.

Because of this “symbol overloading”, if you’re trying to use phonetic transcriptions as instructions for pronouncing English sounds, you can be misled.

Have you heard words like “air”, “near” and “sure” with the schwa sound omitted?

The first thing to understand is that the [r] sound already contains a schwa. When you say [r], your mouth is in the same position as for [ə] – the only difference is that your tongue touches the roof of your mouth at the same time. (This similarity is why [r] is often classified as a semivowel.)

To answer your question, I’d say near is never pronounced with a “clean” schwa in American English. When you say ear, your speech organs first take the position for [i] or [ɪ] (almost closed mouth, tongue in forward position), then they begin changing position (mouth begins to open, tongue moving backwards and curling upwards towards the roof of your mouth). By the time your mouth opens up and your tongue shifts backwards to assume the position for schwa, the tip of your tongue is already on its way to the top of your mouth, resulting in [r] or something between [ə] and [r], which then turns into a real [r]. In other words, I don’t think there is ever a moment where your mouth makes a clean schwa without any r-like quality (curled-up tongue). Does this mean that the schwa is omitted? Depends on your definition of a schwa. Is a “schwa with a hint of r” a schwa or something else?

The same goes for /ɛr/ (air), /ʊr/ (sure), /ər/ (adder), and /ɜ:r/ (fur). The last two normally sound like [ˈædr̩] and [fr̩:], respectively. Note the vertical lines which indicate that the /r/ forms a syllable – it acts as a vowel.

Once again, the lesson here is not to take phonemic transcriptions literally: what looks like a sequence of sounds may actually be one sound.