They disagree on nearly every concrete question: speak immediately or stay silent, flashcards or no flashcards, translate or never translate. And all of them arrived. What does that tell us?
There is a question people rarely ask when reading about someone who speaks ten, fifteen, twenty languages: if all their systems work, why do they contradict each other so completely?
One says speak from day one. Another stays silent for six months and only listens. One builds everything on flashcards. Another calls flashcards a trap. One makes translation the central tool. Another treats translation as the thing to eliminate first. All four have people who followed them to a high level.
This piece looks at what they actually do, where they diverge, where they agree — and why the agreement is the part worth copying.
Lomb was a Hungarian interpreter who worked professionally in sixteen languages and was among the world’s first simultaneous interpreters. Her Így tanulok nyelveket (1970), in English Polyglot: How I Learn Languages, sets out her method.
It is almost annoyingly simple: buy a novel in a language you do not know, and read it. Do not look everything up. Do not study the grammar first. She inferred meaning from context, noted what recurred, and opened a grammar book only once she had enough feel for the language to understand what the grammar book was talking about.
What is striking is that she did this in the 1940s and 1950s, without recordings and without the internet. Her central principle — tolerating not understanding — was later given another name by applied linguistics: extensive reading, which Nakanishi’s meta-analysis found improves both comprehension and reading rate.
Arguelles is known for two techniques he named himself. Shadowing: walking outdoors with an upright posture, repeating a recording aloud half a second behind it. He argues that walking and speaking loudly keep the learner alert and make the mouth do real work.
Scriptorium: read a sentence aloud, write it out by hand while reading each word aloud again, then read back what you wrote. Three channels at once, and deliberately slow.
But the most useful thing about him may not be either technique. It is the logbook: he recorded his study hours per day, per language, for decades. Not to display them, but because the number shows what the feeling conceals.
Lampariello’s method is compact and checkable. Take a passage written by a proficient speaker and translate it into your own language. Leave it a few days. Then translate it back into the target language without looking at the original. Now compare.
The value is in the comparison. Your version is usually not ungrammatical — it is simply different. And those differences mark exactly the places where you chose a correct option that was not the used one. This is precisely what Pawley and Syder called nativelike selection, and the exercise lays it out on paper.
Kaufmann, a former Canadian diplomat and the founder of LingQ, stakes almost everything on large quantities of listening and reading in content he genuinely cares about. He argues against early grammar drilling and against forcing speech before readiness.
His position sits close to Krashen’s input hypothesis: language is acquired when we understand messages, not when we learn rules. That hypothesis has been argued over for four decades and few accept it whole — but the part about needing a great deal of comprehensible input is close to uncontested.
Lewis takes the opposite line: open your mouth on the first day, make a mess of it, and treat embarrassment as part of the exercise rather than an obstacle. His argument is that a learner waiting to feel ready never feels ready.
There is research on his side too. Skill-acquisition theory holds that automaticity comes from practising the actual behaviour. You do not get more fluent at speaking by listening; you get more fluent at speaking by speaking.
Both are half right, and that is not a dodge: input builds comprehension, output builds production, and those are two different abilities.
Early or late speech. Lewis on day one; Kaufmann and the full-immersion school after months.
Flashcards. Some put spaced repetition at the centre, mining every sentence they meet; others regard that as turning language learning into list learning.
Translation. Lampariello makes it the main tool; direct-method schools treat routing through your first language as the fault to be removed.
Grammar. Lomb read grammar books — but late, after the feel was there. Plenty of others skip them entirely.
If any single technique were the cause of success, these contradictions could not exist. So the cause must lie elsewhere.
One: daily contact, with a minimum small enough to survive your worst week. None of them works in bursts. All of them have a non-negotiable floor, and it is lower than you would guess — sometimes fifteen minutes. The point is that it never reaches zero.
Two: large volumes of input, in material they actually wanted to consume. Nobody builds a ten-thousand-word vocabulary out of a textbook. They read novels, listen to podcasts, watch programmes they would have watched in their own language anyway.
Three: tolerance of not understanding. Lomb names it outright; the others practise it without naming it. A learner who stops for every unknown word never reads enough to improve.
Four: they count something. Hours, pages, words, consecutive days. The particular number hardly matters; having one does, because your own sense of your progress is not trustworthy.
Five: speech arrives eventually, and they all found a way. Even the self-described input people end up moving abroad, hiring a tutor, or finding someone to talk to. Nobody only ever listens.
Six: they wrote their system down. This may be the least-noticed thing they share. Each of them articulated what they were doing — as a book, a video series, a blog. Writing it forces the system to be consistent, and turns it into something followable on the day the motivation has gone.
Lýdia Machová, a polyglot who studies other polyglots, gives the tidiest answer after interviewing a great many of them: what they share is not a technique but having found a way to make the process enjoyable. People who enjoy it do a lot of it. People who do a lot of it improve.
Put another way, a system’s real function may not be optimisation but adherence. A mediocre system followed for three years beats a perfect one abandoned after six weeks. That is consistent with what the deliberate-practice literature shows: accumulated volume and regularity matter more than the sophistication of any single session.
It also explains why polyglot advice so often fails to transfer. They are describing what kept them going. For you it may be something else entirely.
The language counts are self-reported and hide a great deal. Speaks twenty languages can mean two at C1, four at B1 and fourteen at the level of greetings. There is nothing wrong with that, but it is not what the number suggests.
There is a commercial layer. Many of these people sell courses, apps and books. That does not make their methods wrong, but it creates an incentive to present a system as a secret rather than as one option among several.
Survivorship bias is real. We hear from the people whose systems worked. We hear nothing from the probably far larger number who followed the same system and quit. You cannot infer cause from a sample of winners.
Choose input you already wanted to consume. If you do not watch the news in your own language, do not build a plan around the news.
Set the minimum you can hit on your worst day. Not your best. Ten minutes that never breaks beats ninety minutes once a week.
Pick exactly one number to count. Not four. One, written where you will see it.
Decide your look-up rule in advance. For instance: only look up a word the third time you meet it. A rule saves you from re-deciding on every page.
Schedule the speaking rather than waiting to feel ready. That feeling does not arrive on its own.
Write it down, and review it monthly. If you have skipped part of it for three weeks, do not blame yourself — fix the system. A rule you cannot follow is a wrong rule, not a wrong learner.
In the end, the most transferable thing about all these people is not any technique. It is that each of them had a system, devised it themselves, and stayed with it long enough for it to work.
The same article, told in plain words — for younger readers, or for anyone who wants the point quickly.
Here is a question people rarely ask when reading about someone who speaks fifteen languages: if all their systems work, why do they contradict each other so completely?
One says speak from day one. Another stays silent for six months and only listens. One builds everything on flashcards. Another calls flashcards a trap. One makes translation the central tool. Another treats translation as the first thing to eliminate.
All four have people who followed them to a high level.
If any single technique were the cause of success, these contradictions could not exist. So the cause must lie elsewhere.
1. Daily contact, with a minimum small enough to survive your worst week. None of them works in bursts. All of them have a non-negotiable floor, and it is lower than you would guess — sometimes fifteen minutes. The point is that it never reaches zero.
2. Large volumes of input, in material they actually wanted to consume. Nobody builds a ten-thousand-word vocabulary out of a textbook. They read novels and watch programmes they would have watched in their own language anyway.
3. Tolerance of not understanding. A learner who stops for every unknown word never reads enough to improve.
4. They count something. Hours, pages, consecutive days. The particular number hardly matters; having one does, because your own sense of your progress is not trustworthy.
5. Speech arrives eventually, and they all found a way. Even the self-described input people end up moving abroad, hiring a tutor, or finding someone to talk to. Nobody only ever listens.
6. They wrote their system down. The least-noticed thing they share. Writing it forces the system to be consistent, and turns it into something followable on the day the motivation has gone.
A polyglot who has interviewed a great many other polyglots gives the tidiest answer: what they share is not a technique but having found a way to make the process enjoyable.
People who enjoy it do a lot of it. People who do a lot of it improve.
Put another way: a system's real function may not be optimisation but adherence. A mediocre system followed for three years beats an optimal one abandoned in a month.
In the end, the most transferable thing about all these people is not any technique. It is that each of them had a system, devised it themselves, and stayed with it long enough for it to work.
These articles summarize well-established research in learning science and linguistics. Key sources and further reading:
The question assumes focus is a muscle you train. The strongest evidence says the training barely carries over, and the...
No time, no talent, too old — those are explanations, not causes. What can be measured is duller: one sentence with a ti...
Not one event but four, running on clocks a billion-fold apart: a wave at 400 milliseconds, a word locked in overnight,...
