2026-04-10
10 分钟Today I want to talk about flashcards versus massive input
and what have I learned about why one works or the other works.
Flashcards have quite a history.
Even in the 17th century there was a famous educator called John Amos Comenius who developed a form of flashcard
for connecting words to pictures to be used in education.
John Stuart Mill apparently used flashcards to learn Greek.
With the Greek word on one side and the English word on the other.
And then the flashcards moved to a form of spaced repetition system.
And in the 19th century, Ebbinghaus did some research on the memory curve and how soon after first encountering an item,
a word or whatever it might be, one should ideally meet that again and again.
And he developed this ideal frequency to relearn something before we forget it.
And in 1972, a German called Leitner wrote a book called So lernt man lernen,
describing how, you know, learning was like a filing cabinet.
And I remember reading the book and I thought it was very interesting.
And it introduced, again, this idea of spaced repetition according to a certain formula.
And this was then taken further in the late,
you know, the 1990s with algorithms and Wozniak of SuperMemo and Anki and so forth.
So it's very much a mainstream part of language learning.
But another mainstream part of language learning is compelling input, massive input.
And I want to look at these two based on my recent discovery of the thoughts of Jeffrey Hinton,