James Murray - HOW TO LEARN LANGUAGES: From the latest research to a step by step plan

ACKNOWLEDGEMENTS

Firstly, a big thank you to the countless friends and strangers who have patiently l istened to us try to talk their language over the years. We thank the numerous inspiring polyglots who have shared their secrets and techniques with us and made the world a more understandable place. Finally we thank the friends and family that have made this handbook possible: Fiona Parr, Chris Chillton, Sam Caldwell, and Joachim Meuli.

Chapter 02

HOW WE HEAR SOUNDS

We saw that the first building block of a language is the set of sounds (or phonemes) that make up speech. Being able to recognise and reproduce these sounds will allow us to start feeding our language learning engine.

The worlds languages contain roughly 800 sounds (six hundred consonants and two hundred vowels). On average a single language will be made up from around 30-50 sounds. At the extremes of this scale you have Pirah in the Amazon with only 11 sounds, and !X, a dialect spoken in Namibia, with 141!

We are born with the ability to hear all these sounds - which makes life very confusing for babies who hear all the subtle variations in the pronunciation of words around them. By about six months to one year of age we have learned to filter out all those subtle differences so we only hear the 50 or so that make up our language. This makes things a lot easier. Unfortunately it is this that creates the first major barrier to learning your new language. You literally cannot hear the different sounds that make it up!

Most phonemes are quite subtle variations of familiar sounds. Korean has a lot of consonant variations; for instance, g, k, and kk (, , ) which can all sound like k to an English-speaker and be difficult to pronounce distinctly.

In a study comparing Japanese and Americans, researchers used brain scans to see whether an individual could hear the difference between two given sounds. When you play an American a recording of rock rock rock rock lock in a brain scanner, there will be a sudden increase in brain activity when lock is pronounced.

However for a Japanese adult there will be no significant activity. This is not so for a Japanese baby, who can still recognise the different sounds until between 6 and 12 months of age.

We group varied sounds into specific groups based on the context in which we hear them. Using the language changes our perception and makes us deaf to some sounds and more attuned to others. As we hear different sounds we group them together based on what is most likely. If you listen to all the r and l sounds in a typical day in America you would create a map as follows:

In a Japanese household that same graph would look something like this:

So, the first thing we need to do when learning a language is to reactivate the sounds that our brain has decided we dont need to perceive.

In later studies, the same rock...rock...rock...lock sequence was played to Japanese adults. They were asked to press one button when they heard lock and another when they heard rock. The results were predictably terrible.

With a second group of participants, feedback was given after every correct guess. Once the incentive was introduced they began to learn and the performance increased. When the same brain scan was re-performed the distinction between rock and lock was significant.

This research is the foundation behind a technique called minimal base pair practice developed by the team at Fluent Forever that allows you to rapidly learn the new sounds needed for a given language. You take two words (the base pair) and you listen to a recording and guess which one it is and then get immediate feedback. In English an example of some pairs are:

• Rock and Lock
• Moss and Mouse
• SUS-pect and sus-PECT

Why its important to make sure you can hear all the sounds.

If we learn the sounds first, we make all subsequent study of our language more efficient. Words will sound more familiar so we will be able to hear them directly and therefore remember them more quickly. But this isnt the only advantage. Having focussed on the sounds, we will be able to hear subtle variations in those sounds, which will allow us to recognise sound rules and exceptions to those rules.

Sound rules give you the basis for knowing which sounds can be combined. Our internal language machine is primed to recognise and copy these rules and will start to unconsciously internalise them early on.

There is a common test performed on young children that demonstrates this. Researchers show children a picture of a strange bird and say: This is a wug!. Then they show the children two of them and say, Now there are two of them! There are two and the kids gleefully exclaim, Wugs! pronouncing it wugZ.

While this appears very simple, there is a subtle linguistic rule at play. They know that the plural of this word they have never seen before is pronounced /z/ and not /s/. By developing our ear we will internalise these rules far more quickly.

SUMMARY

Hearing new sounds requires us to rewire our brain. It is a physical skill and needs to be practised and trained. To do this we can listen to minimal base pair recordings. It is good to spend time learning these new sounds early on in our language learning journey since it will improve the efficiency of memorising sounds and sound rules. Remember there will only be a handful of new sounds that you will need to learn.

PRACTICE FRAMEWORK

• Input: Listen to pairs of words (minimal base pairs)
• Output: Differentiate between the words correctly, pronounce the words correctly
• Feedback: Flashcard gives correct answer, native speaker confirms your pronunciation

Fluent Forever is an app that helps you to do this.

Chapter 01

WHAT IS A LANGUAGE MADE FROM?

To understand what we need to learn, it is helpful to understand what constitutes a language.

On the surface, a language is composed of four concrete components:

• Hearing sounds (Listening)
• Making sounds (Speaking)
• Reading (or occasionally feeling) symbols and letters (Reading) Using symbols and letters to make words and sentences. (Writing)

We need to ensure we can hear all the sounds used to communicate our new language, and once we can hear them we need to learn how to make those sounds. These first two components are generally the most problematic for getting started.

The good news here is that the list of sounds and symbols in a language is (generally speaking) very limited! For example, there are only 44 different sounds (phonemes) in English, and only 33 letters in the Cyrilic (Russian) alphabet.

In addition, since many sounds are shared between languages, you will only need to learn a handful of new ones. The important thing to note at this stage is that hearing and creating new sounds is 99% physical - involving the lips, tongue, palate, teeth and facial muscles.

These sounds and symbols are then used to communicate the abstract, sense-making components. In this respect a language consists of two things:

• A finite vocabulary
• A finite grammar