Study Alone or With a Friend? Two Ways to Learn Something Hard

For most hard material, the best answer is both, in a particular order: alone first, then with a friend. If you can only have one, it depends, and here is on what. It depends on whether your trouble is paying attention or noticing what you don't yet understand. Solitude helps with the first. Company helps with the second.

I tried both before writing this, with a cup of tea and a cheap notebook, as usual. The sample was small and the results were mine alone, so please treat what follows as a set of observations to test, not a verdict.

What studying alone gives you

The first gift of solitary study is focus. Hard material places a heavy load on working memory, which is the small mental workspace where you hold a few things at once while you connect them. Conversation, however pleasant, spends some of that workspace on the other person: their face, their pace, what to say next. Alone, all of it goes to the problem.

The second gift is your own pace. Hard ideas rarely arrive at an even speed. You may fly through three pages and then need forty minutes on one paragraph. Alone, you can stay stuck on that paragraph without anyone watching the clock or politely wanting to move on.

The third is privacy for being wrong. You can write a bad guess, cross it out and try again, and nobody has seen it. For some people that freedom is the whole point.

Solitude has costs, and they are quiet ones. The main one is what psychologists call the illusion of explanatory depth. Most of us feel we understand how a zip or a bicycle gear works, until we are asked to explain it step by step and find a gap. Rereading a page produces the same feeling. The words are familiar, so the ideas feel familiar, and familiarity is easy to mistake for understanding. Alone, there is nobody to ask the awkward question, and the other cost follows from that. Without a deadline, "I'll look at it properly tomorrow" can run for a surprisingly long time. I speak from experience, and from a notebook with a good many blank pages.

What a study partner gives you

A partner gives you three things: questions, encouragement and a gentle deadline.

Questions. A good partner doesn't need to know the subject. They only need to be curious enough to ask "but why?" at the right moment. Questions find the gaps that rereading hides, because you cannot answer them by recognising a familiar sentence. You have to produce the answer, and producing an answer from memory is a stronger form of practice than looking it over again. That much is well established in learning research.

Encouragement. Hard things feel worse alone. When a friend says "I couldn't follow that bit either", the difficulty becomes ordinary, and ordinary difficulties are much easier to keep working at. It is a small kindness with a practical effect.

A gentle deadline. "Thursday at ten, at the café" is firmer than "sometime this week". Most of us will let ourselves down before we let a friend down, which is not noble but is useful.

The costs here are real too. Sessions can drift into talk about everything but the subject. Paces can differ, so one person is bored while the other is lost. And there is a subtler trap: a partner who nods along can make you feel you have explained something well when you have only explained it fluently. They may be filling the gaps in your explanation from their own head, kindly and without noticing.

A worked example: why do we have seasons?

Here is the gap-finding at work. Ask yourself why it is hot in Australia in January, and give your answer aloud, not silently.

Many people begin: "Because the Earth is closer to the Sun at that time of year." It sounds right, and the explanation feels complete. A partner can find the problem with one question: "If it's about distance, why is it winter in Canada in January while we're swimming?" Distance would affect the whole planet at once.

Alone, you might never ask that, because the first answer arrives with a sense of finished business. Once the question is asked, the better explanation comes into reach. The Earth's axis is tilted, by roughly 23.5 degrees. For part of the year the southern half leans towards the Sun, so sunlight hits it more directly and the days are longer. Six months later the northern half leans in instead. In fact, the Earth is slightly closest to the Sun in early January, by a few per cent of its average distance, and that difference is far too small to explain the seasons.

Notice what happened. Nobody taught anything. A question exposed a gap, and the explanation was then rebuilt, which is the job a partner does best. Programmers have a joke-that-isn't-quite-a-joke about this. Some explain their stuck code, line by line, to a rubber duck on the desk, and find the bug halfway through. The duck says nothing. The act of explaining does the work. A human partner is a duck that can also ask follow-up questions.

Seneca, in one of his letters, passes on a saying that people learn as they teach. I find it reassuring that this was already an old observation.

How to make a pair session useful

A pair session without a shape tends to become a pleasant chat. This is the format that worked best for me. Adjust it freely.

1. Prepare alone first. Each of you spends 20 to 30 minutes with the same section of material, separately. This uses solitude for what it does best: quiet, unhurried first contact. It also means neither of you arrives empty-handed.
2. Put the notes away. Closed books are the point. Explaining from memory shows you what you actually hold, rather than what you can recognise on the page.
3. Take turns explaining. One person explains a single idea for about five minutes, in plain words, as if to someone new to it. Then swap. Keep to one idea each, so it stays manageable.
4. Let the listener ask, not correct. The listener's job is one or two questions, such as "why does that happen?" or "what would change if…?", not a verdict. Correcting straight away takes the discovery away from the explainer.
5. Mark the hesitations. Whenever the explainer slowed down, said "sort of" or waved a hand, put a small tick in the margin. Those ticks are your study list.
6. Finish apart. Spend a short stretch alone, going back to the ticked places. This again uses solitude for what it does well, which is slow, private repair.
7. Book the next one before you leave. The gentle deadline only works if it exists.

Two further tips, both learned by getting them wrong. First, two people is usually enough. With three or four, the quietest person tends to become the audience. Second, a partner who is slightly ahead of you or slightly behind you can both work, as long as you take turns being the explainer. Nobody should be cast as the teacher for the whole session.

When one way suits you better than the other

Some people thrive on one approach and wilt on the other, and neither is a flaw.

Someone who finds conversation tiring, or who thinks best in long silences, may leave a pair session emptier than they arrived. A person who needs to be heard to think, or who drifts without an appointment, may sit alone with a book for an hour and retain less than they would from fifteen minutes of explaining aloud. Often it comes down to temperament, though it can also depend on the day and on the material. Memorising a list of terms may suit solitude, while untangling a concept often suits company.

If you aren't sure which you are, here is a small experiment:

1. Pick two sections of similar difficulty from the same subject.
2. Study one alone and the other with a partner, using the format above.
3. A few days later, without looking at anything, write down what you remember of each, and then check it.
4. Notice also how you felt: energised, flat, anxious, pleased.

Both results count. If you remember more from one but dreaded it, that is worth knowing too, because the method you will actually keep doing often beats the one that looks best on paper. Remember that a single trial is a weak test. I've been surprised by my own results more than once, and I am still learning which of my hunches to trust.

If a friend isn't available, you can borrow some of the effect: explain the idea aloud to a houseplant, a pet or a rubber duck, or write it out as a letter to someone who knows nothing about it. The result isn't quite the same, since a plant asks no questions, but it does catch some gaps.

A question to carry around

What puzzles me is why speaking an idea aloud shows up gaps that silently thinking it through does not. Part of the answer is probably that speech has to be complete, in order, and in full sentences, while thought can quietly skip ahead. I don't fully know, and I'd like to learn more. If you try the experiment above, I'd be curious which way it fell for you, and whether the answer changed with the subject.

Thank you for reading, and good luck with the hard thing. It is, I'm told, hard for nearly everyone.