Handwriting is better than typing when learning a new language, study finds

You have my upvote. "Learning languages" by what people most often mean by the phrase is very different and distinct activity than learning letters. Yeah, you need letters to read and write, but most of the learning required to learn a language is spent almost entirely in a different modality.

This is exactly true from my experience. When you learn an new writing system, seemingly insignificant differences can have a big impact, and writing by hand forces you to confront them.

Consider in Latin the difference between an I and an l, if you have never seen the latin alphabet before this might be quite confusing. Many writing systems have much more subtle differences than this, especially in cursive forms. e.g. if you use a keyboard you will never learn that م and ه are easily confused in Arabic if you don't write them properly.

When learning vocabulary you want to focus on throughput. In the long term it does not really matter if you get a letter wrong here or there, it will correct with enough exposure and practice. The important thing is to do as many drills as possible with your chosen space repetition tool. A keyboard is a better tool than a pen for that.

Anyone who has shown a kid an iPhone would know that touch screens are immensely more intuitive for kids. They can write words in days compared to months of rote handwriting practice. Of course the press around this study will be flawed. And of course this study didn’t have a control, like learning to play the violin or a similar motor skill.

Wow. Yeah, I never made this connection before.

I type you instead of your frequently, mix up homophones like your/you're and there/their/they're etc. reasonably frequently. I very rarely make these mistakes in hand writing OR on my iPhone; only at the keyboard.

I wonder if this has to do with how fast typing is. Is there similarly diminished learning in very slow typers? Does a shorthand expert fail to learn as much?

This is an original thought to me.

I wonder how grammar evolves now that most language is typed rather than handwritten.

I think there’s a certain “muscle memory” involved. When you’re writing a word by hand you have to think about each letter, however briefly, as you write it: “w-o-r-d”. But when you’re typing a word you don’t think about each letter so much as you think about the pattern your fingers need to make, if that makes sense?

It’s like when someone asks what the keyboard shortcut is to take a screenshot. Uh, I don’t know? I use it all the time and I couldn’t tell you if I’m hitting control or command or option because the underlying inputs don’t really matter. I only need to know the pattern my fingers are supposed to make and so that’s all that gets committed to memory.

Less fluency with handwriting is a possible explanation. If you did it a lot more, you might run into muscle memory mistakes.

fun thing, it's other way around for me. I hate writing, because I skip letters all the time, have to cross a letter or two to correct that mistake, and make a mess as a result. Not much practice for more than a decade, of course. I have no such problem when typing.

Additionally (and mainly?) a skill learnt as a young child - eg. handwriting - is presumably more likely to be fully absorbed and "foolproof" than one like typing which at least in former generations (so may not apply in your case) was learnt later (and often to a lesser competence) as a teenager or adult.

I made tons of mistakes when writing. It's the reason there's an eraser on the back of the pencil and lots of larger erasers as well as things like white-out

> Writing... is not that. Yes, you can rush. But it's slower, and I find that it helps me stay on single lines of thought. There's no manic jumping around, edit here, edit there. You can write the words, slowly even, and ruminate in their meaning.

I really like what you’ve said here. This rings true for me as well. For years I was on a hunt for a better note taking app. I tried everything from a basic Mac notes app to Sublime Text to more niche tools like Omnifocus. Every time I ended up writing out the most critical notes by hand in a spiral notebook because it was more effective for me personally to retain that information.

OTOH, of you do need to keep up (with a lecture, for example), handwriting hurts because you don't have time to finish.

I don't think the speed matters as much as that drawing words cements them more solidly in the mind than touching them. So more an imbalance in process difficulty than speed.

The same effect happens in note-taking. I almost never review my handwritten notes, but the act of writing them is probably why I don't have to review them. When I'm recalling information, I frequently remember where and how on the piece of paper I wrote it.

iA Writer in typewriter focus mode is great for me this way. It encourages me to treat a document as append-only until I'm completely finished writing it.

Gives side-eye in fountain pen.

I really do prefer fountain pen on nice paper, but there are times where you just need software.

Followup: I realize I'm replying more to other comments in this HN thread about handwriting notes being superior to typing them, which isn't at all what the article's about.

The article said handwriting is better for learning a language, and that seems to intuitively make sense. It didn't talk about note taking in general.

The thing is, that returning to your notes is the least important part of taking notes. What writing notes longhand does is do a better job of encoding what you're taking notes about into your long-term memory.

Anecdata: As part of an ongoing project,¹ I spent about two years teaching myself Biblical Greek. As part of this project, I ended up filling a couple notebooks with handwritten notes—writing out tables of conjugations for every verb that was taught, handwriting translations to and from Greek, etc. What I found was that even though I never actually went back to the previous notes, I had good recall of the information that I practiced in this fashion. I think (but what do I know?) that part of it was the spatial component to the note-taking. I can't find stuff in code that I've written without searching,² but I could go back to my handwritten notes from learning Greek (2017–2019) and find, say, what I had written about the aorist passive without much difficulty.³

---

1. https://www.dahosek.com/category/dewey-decimal-project/

2. You might think, “big deal—then search” but what if the search is not by some thing easily turned into text but rather some concept which I don't remember exactly what I typed to do it, or the searchable thing ends up being some bit of text that recurs so frequently that the search is effectively useless. I have a much easier time finding things I've read on paper from my personal library of around 1200 printed books than I do finding something I've read online even with the fast searching capabilities of the internet (although I will admit that it's a lot easier to find disposable writing again online than in print).

3. That spatial context helps in other ways too—I can find things from grad school notes (2002–6) pretty easily as well and there was one time in 1999 that I pulled up a quote from a book that I'd read in 1992 (which had no index) in just a couple minutes thanks to my memory of roughly where in the book and where on a page spread the passage appeared.⁴

4. Yes, I know that, especially with that last anecdote that I'm getting into stuff that's likely well outside the norm, but the basic point that spatial context helps with memory is something that's well-studied and documented. We don't get that with typing on a screen.

I know what you mean about writing causing pain in the hands and wrists...

Have you ever tried cursive? I can write much longer and with less pain when using cursive - really night and day.

> shackled with vast tooling that lets me selectively encrypt sensitive parts, sync them across all my devices, instantly search through everything, turn notes into reminders with a tap of a button

I know that this is subjective, but I thought that this mattered, but ultimately it did not. I rarely (/never) refer back to my electronic notes but I do occasionally browse through my handwritten notes. The advanced features that you mentioned were cool to play with, but are a distraction from the main purpose of note taking.

For what it's worth, I can barely make reminder apps/TODO lists work for me, maybe I should try a hobonichi or something. (/me makes a note to buy one, which I will probably never get around to).

> I wish they would do studies like this with children / teenagers.

They have. I don't have said studies in front of me. But the gist of it seems to be that typing is a quicker lower process skill (pushing a button), where as handwriting is a slower higher intensive process, (thinking about the letters you write).

My opinion is that physical note taking, when not transcribing, requires you to process the information you are given, then summarizing it in your own words. With the benefit of handwriting being that you have more time to process the information.

I wonder this too - I know that I learn much more rapidly and the knowledge "sticks" when I handwrite notes but just anecdotally, I feel like the effect is "lessened" when I handwrite digital notes.

When I learned Japanese, I realised after a while I could recognise a few kanji, but could visualise barely any of them. As in, I could use the language already, but if you asked me to write some hiragana character, I'd be totally stuck.

Practicing writing, even with apps made for that purpose helped a lot.

Letters change shape depending on where they are in the word, so almost certainly! :)

I wish the same would exist for reading. One sentence at a time.

im thinking about note taking programs that have a distraction-free mode or append-only mode... maybe a slow mode where you can only type as fast as you can write would help somewhat

It doesn't transfer as easily to other people but mindmaps or lukasa (for an ancient predecessor) are infinitely information dense and naturally built to the particularities of their builder's mind.

>This seems to be the input-level dual of the meme that reading from physical paper leads to greater retention than from a computer screen. I'm curious as to what the cause could be for this one, as it seems less intuitive than the writing case.

One of the possible reasons would be that we consume text content on paper differently than on the screen.

Another reason would be that when you read text on paper you engage with a unique physical object. Engaging multiple senses and adding more cues/anchors helps with creating stronger connections in your brain and improves memory retention.

That's also probably why writing things down with a pen helps you remember more and for longer--more parts of your brain are involved in the process of note taking.

The physical act of writing engages the brain more than typing on a keyboard does. I claim that the physical motion and having to form shapes, as well as the more tangible spatial procession of your physical movement across the page reinforces the memory through association in a way that just pushing a key does not. Your hand is very intimately involved in making those marks on paper. Pushing buttons causes things to appear on a screen, but there is little tactile engagement and it isn't very differentiated (all pushes of a key are approximately the same). Writing is also more deliberate, whereas you can type while half asleep. The quality of your writing also depends on effort. Typed letters appear the same no matter how you press the key. You are mainly concerned about pushing the correct key, but how to press it is unimportant and there isn't much room for variation anyway.

We are in a field where we get to experience these ideas head first.

For instance I remember having to learn how git works, and followed online videos on the base concepts, tutorials with commands to type etc. Most importantly there was no paper nor handwriting involved at any point.

I personally don’t feel like it was holding me down and it would have been a royal pain to introduce a non digital media in the middle of it.

I’d wagger I’m not the only one, and the majority of us got to terms with digital learning and how to be efficient with it.

The other extreme, like at my university, is to drill writing everything (and cover less-interesting topics like telling the time, and almost useless words like '雙胞胎' - how many twins are there in China, Confucius Institute?!)

I suspect most of your problem comes from trying to read isolated characters.

Maybe in the short term, and to learn the basics/conversational Chinese, however I believe in order to fully and best understand the writing system it's still best to learn how to write the characters. Plus, the deliberate act of creating the characters stroke-by-stroke seems to better implant them in my brain more than just typing out some pinyin, but maybe that's just me. Spending 5-10 minutes writing a character helps me remember the meaning and shape of it much better than an hour doing pinyin/duolingo esque practice.

But I still probably can never see myself memorizing some of the more complex characters that even native Chinese folks have problems with :P

> Typing Chinese is so much faster and easier than writing by hand.

The article is talking about how learning is facilitated by handwriting in a way that typing does not, and not about which results in more quickly produced text. (Also, it should be clear that how well you know the language will also put a bound on how fast you type.)

None of this is news. I've had teachers tell me this years ago. Writing engages the brain in a way that typing keys does not. That a study corroborates this claim is not surprising to me.

I'm learning Japanese, and I'm learning the kanji using Heisig's Remembering The Kanji method. I combine that with flashcards on Anki on my PC. When it shows a keyword card, I'll write out the character on paper or I'll trace it in the air with my finger, then check if I got it right. That works surprisingly well to memorize them.

That said, the memory pathway should work for reading and writing, that is, recognition and production. I need a further skillset (pronunciation) in order to do that production on a computer, since being able to write individual characters by hand without any knowledge of the words or pronunciation isn't very useful.

I'm a second-generation Chinese-American who learned my first ~200 Chinese characters by handwriting and then everything else from the internet.

I have no idea how to handwrite half the characters I can recognize, read, and "write" now thanks to typing with a Pinyin or Jyutping keyboard.

I find though that it's impossible to read cursive (hand-written) Chinese unless you've written a lot yourself (with the correct stroke order). Read printed Chinese and type using pinyin, sure. Read hand-written Chinese - no way.

(n=1, too)

Heh, its just a paper reporting results that they found. There is nothing really to disagree, as it is not an opinion. I think science reporting should provide this context rather than produce some kind of headline that is treated as The Truth :)

Anyway, when I was trying to learn French, I found that for myself no single method worked. I learnt by reading, listening, writing, conversing, singing some random french song, etc. My belief is that learning is multimodal, and there is support for this view in literature. I think these types of studies are still valuable as they control for one variable and provide some insight on that. I'm sure we will slowly converge on the learning model most likely to promote long term retention.

I agree that copying characters in Chinese made my hand hurt, and I didn't internalise any of the meanings.

https://pingtype.github.io/docs/failed.html#copyingCharacter...

Reading parallel text processed with Pingtype was far more useful for me.

1. There's more to determining if an effect is in the statistical noise than just sample size. Effect size, for example, also comes into play.

2. Not every study will, or even should, try to be the final word on some matter. If you come up with some novel hypothesis in social science, you're not going to jump into a 2,000 person study.

It might be noise, but you can't tell from N=42 alone.

It is possible to find "significant" results with small groups (my back-of-the-envelope calculation suggests that the proportion of success between the groups must differ by more than 30 percentage points to be p<0.05). They might well have done the maths and might have highly significant p-values.

Note the supplementary materials [1]. I don't understand them without the paper, but they sure do have a lot of "*** p < 0.001" results.

(Entirely unrelated side note: A shame that Sci-Hub currently does not add papers due to litigation...)

[1] https://journals.sagepub.com/doi/suppl/10.1177/0956797621993...

42 happens to be the answer to the "Universal Question About Life, The Universe and Everything". [0]

[0] https://en.wikipedia.org/wiki/The_Hitchhiker%27s_Guide_to_th...

PS: I get it that some people will down-vote this because jokes like this do not really stimulate an intellectual discussion. But reading something by Douglas Adams is a really good idea.

This is a great question.