The familiar advice is to mix everything up. But for plain vocabulary, studying in blocks comes out ahead. What decides it is not whether you mix, but whether the two things you are mixing are ones you confuse.
You have three things to practise and one hour. Arrangement one: all of A, then all of B, then all of C. Arrangement two: A, B, C, then A, C, B, and so on, mixed. The same number of items, the same number of minutes. Only the order differs.
The first is called blocked practice, the second interleaved practice. It sounds like a trivial detail, but it is one of the few interventions that has been measured over and over across dozens of studies.
And how it feels is not in doubt: blocked practice feels far better. Do twenty items of the same pattern in a row and you can watch yourself improve. Mix them and you stumble more, work slower, and come away thinking the session went badly.
Then the test comes a few days later and the order reverses. Pooled across studies, interleaving beats blocking by a moderate margin — clear enough to see, but nothing like night and day. The feeling of learning well during practice turns out to be a poor guide.
The usual explanation is that mixing is harder, and harder is better. That sounds right but it is far too vague, and it leads people to the wrong conclusion — that everything should be mixed.
The explanation that holds up is much more specific: interleaving works because it puts two confusable things side by side, close enough for you to see what separates them. Blocked practice never shows you that contrast. Twenty items of one pattern in a row ask you to discriminate nothing.
There is evidence that this is not simply spacing in disguise. When filler items were added so that the time between encounters was equal in both conditions, the interleaved group still won. And spacing that broke up the juxtaposition actually hurt, even with total spacing held constant.
Put briefly: interleaving is not mixing for the sake of mixing. It is placing next to each other exactly the things you get wrong.
Once the mechanism is clear, it is obvious why mixing does not always pay. The moderators have been measured and they are specific: interleaving earns its keep when the categories resemble one another, when items inside a category differ, and when the material is complex.
And here is the exception that matters most to language learners. For plain vocabulary, blocking wins — by roughly the same margin as above, only in the opposite direction. The reason fits the mechanism exactly: two new words are not two categories you need to tell apart. They are simply two things not yet in your head. Mixing them adds load and offers no contrast in return.
For grammar the finding is not a winner but a trade-off. During practice, the blocked group used the structure more fluently; the interleaved group used it more accurately but more slowly. You choose by what you currently need, and there is no general answer.
Beginners are clearer still. With little prior knowledge, blocking first is better. Difficulty only becomes useful once you have something to discriminate with. Mix before you can recognise the pieces and all that is left is the difficulty.
So the rule is not to mix everything. It is tighter than that: mix what you confuse, and block what you are still installing.
Worth mixing are the pairs built to trip you up: 了 and 过 in Chinese; ser and estar in Spanish; Perfekt and Präteritum in German; 은/는 against 이/가 in Korean; the second and third tones when you drill pronunciation. Practised separately, each is easy. Side by side, the gap you have not actually closed shows itself.
Worth blocking are brand-new words, a script you have only just started, a sound you cannot yet produce. None of these needs to be told apart from anything — they need to be installed first.
The sequence follows naturally: block until you can use the thing at all, then mix it with whatever it keeps getting confused with. And expect to feel worse while you do it. That is the price, and it is the same illusion that spacing your reviews produces.
The same article, told in plain words — for younger readers, or for anyone who wants the point quickly.
Think of two people practising for the same hour. One does twenty of exercise A, then twenty of B, then twenty of C. The other jumps about: A, B, C, A, C, B. Same exercises. Same hour. Only the order is different.
The one doing them in blocks feels good. You can watch yourself getting better. The one jumping about keeps stumbling and goes home thinking it was a bad session.
A few days later there is a test. The one who jumped about does better.
No. And this is where most advice stops too early.
Mixing does not work because it is harder. It works for a very particular reason: it puts two things you get confused side by side, close enough that you can see the difference.
Twenty of the same exercise in a row never shows you a difference. There is nothing to tell apart. That is the whole problem with it.
You might think the benefit is simply leaving gaps between attempts. Researchers checked. They made the gaps equal in both groups, and the mixed group still won.
They also tried breaking up the pairs while keeping the total gap the same. That made things worse. So it is the being next to each other that matters, not the gap.
Here is the part that surprises people. For learning plain new words, blocks are better.
It fits the reason above. Two brand-new words are not two things you keep confusing. They are just two things you do not know yet. Mixing them adds effort and gives you nothing back.
Same for beginners. If you cannot recognise the pieces yet, mixing only leaves you with the difficulty and none of the benefit. Do it in blocks first.
The rule is short: mix what you confuse. Block what you are still putting in.
Mix these, because they exist to trick you:
Block these, because they are not competing with anything yet:
So: work in blocks until you can use the thing at all. Then mix it with whatever you keep confusing it with. And expect to feel worse while you do — that feeling is the price, not a warning.
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,...
.jpg){kind=link}
_LCCN2016824250.jpg){kind=link}
.jpg){kind=link}