Back to Marginota

Why reading works

You don't learn a language by studying it. You learn it by reading.

That's the bet behind Marginota. It isn't a hunch — it's one of the most replicated findings in second-language acquisition research. Here's the short version.

Comprehensible input

In the 1970s and 80s, linguist Stephen Krashen proposed that we acquire languages much the way children do: by understanding messages, not by memorizing rules. His Input Hypothesis says acquisition happens when a learner reads or hears language they can mostly understand, pitched just slightly beyond their current level — “i + 1.” Consciously studied grammar, he argued, doesn't turn into fluent ability.

Reading beats drills in the classroom

Elley and Mangubhai (1983) ran the famous “Book Flood” in Fiji: primary classrooms were flooded with high-interest storybooks and pupils read every day, while control schools kept the standard audio-lingual drill program. The reading groups outperformed the drill groups on reading comprehension and other language measures — and the advantage grew even larger in the second year, spilling over into exam results in other subjects, including mathematics.

An honest caveat

Krashen's strongest claim — that comprehensible input is the only thing that matters — is debated. Later research, including Long's Interaction Hypothesis, shows that conversation and producing language help too. What isn't debated: large amounts of understandable reading reliably build vocabulary, grammar instincts, and fluency. We build on the consensus, not the controversy.

How Marginota applies it

Start reading →