There’s a scene in the movie Love Actually where Jamie, played by Colin Firth, is learning Portuguese. He’s sitting in a classroom with row after row of other language students listening to headphones and repeating simple Portuguese phrases, over and over again.
You might recognise the language learning trend that this scene was referencing. It is called the “Audiolingual Method” and became popular throughout the 1940s through the 1960s, declining after that. The idea was that if you heard something enough, and you repeated it, you could memorise it and eventually learn the language.
That is just one of probably hundreds of language learning theories that have picked up steam at some point in the last century and then faded away.
There are many others.
When looking at the wide variety of approaches to learning languages, you might be tempted to ask, “Do we actually know anything about how people learn languages?” Especially when so many websites and services claim that their method is “based on science!”
It turns out that we do know quite a bit about language learning, and one of the concepts that has particularly strong support in the research is the input hypothesis developed by the linguist Stephen Krashen.
So, let’s dive into that.
In this article, I want to outline the input hypothesis and describe what it proposes about how we learn language. You’ll learn that, if you want to learn English, you will make progress fastest by ensuring that you create opportunities to expose yourself to comprehensible input in English.
What is the Input Hypothesis? The Input Hypothesis was developed by Stephen Krashen in the 1970s and 1980s. It’s actually a group of 5 hypotheses. They’re a bit complex, but here’s a very simplified version of them:
The Input Hypothesis states that language learners improve in a language when they are given language input that is slightly more advanced than their current level. Krashen called this “i + 1” where “i” is a person’s current language level and “+1” represents language that is slightly more advanced than their current level. The Acquisition–Learning hypothesis states that acquisition of language is different to learning language. Krashen argued that learning—what we do when we study grammar rules—doesn’t work nearly as well. Instead, language is acquired and that happens through an unconscious process when we are exposed to comprehensible input. The Monitor Hypothesis states that consciously learning language (like studying grammar rules or doing vocabulary exercises) can help a person monitor language output, but it doesn’t result in improvements to using language. In other words, learning grammar rules can help you measure your language ability, but not really improve it. The Natural Order hypothesis states that language acquisition happens in a natural order, which is pretty much the same for everyone. It further says that language instruction doesn’t change this “natural” order. The Affective Filter hypothesis states that affect—how you’re feeling—changes language acquisition ability. Krashen argues that negative emotions, like embarrassment or fear, make a person less able to acquire a language. What does all that mean? That’s all a bit complex, but, very simply, Krashen is saying this: the process of “learning a language” is not the same kind of process as, say, learning geography or philosophy. We can’t read a book about it and then come to “know” it.
Instead, language acquisition happens through an unconscious process. The necessary ingredient—the critical, essential core—of that unconscious process is comprehensible input.
What is comprehensible input? Comprehensible input in English is English language that you can understand. Language inputs are things that you hear (like podcasts, the radio, conversations, and so on) as well as things you read (like books, articles, English blog articles, etc).
Krashen is careful to specify that you can’t just read or listen to anything and improve your language. You have to read or listen to things you can understand. Language acquisition happens best, he says, when the input is just slightly more advanced than your own level.
