Learning to Output
- Three Aspects of Natural Output
- Practical activities to improve your Output
There comes a point to any “immersion heavy” language learner where they have to dip their feet into the scary land of Output. While input is more or less a “solved” problem, output seems to still be a very hot topic among language learning communities because it revolves around so many different variables, and touches on so many different emotional aspects and vibes that each student has with the language and its culture. It is hard to make general statements that will apply to everyone. This is a journey that each person has to go through at their own pace and to their own comfort level.
But first, let me define what I actually mean when I use the word output throughout this page:
- Output is the ability to put together words and sentences in order to convey a message in the most natural way possible.
Let me be clear, this page on output does not touch on the phonetic aspects of the language. It is not going to talk about pitch accent (see:), intonation, diction, etc. While those are all important parts of the language, they are not related to the actual ability to put together a sentence. They can be studied separately and there’s many other guides and pronunciation coaches out there that can help you achieve that. This is not the purpose of this page.
There can be multiple ways to output, for different kinds of platforms, audiences, and styles. Writing a chat message is output. Talking with a friend is output. Making a youtube video monologue is output. Streaming on Twitch to 0 viewers is output. Writing a blog is output. Tweeting is output. Both text and voice, synchronous or asynchronous, count as output. Each of these things have different aspects that can make them harder or easier, but they all effectively boil down to one core skill: your ability to manipulate constructs and phrases of the language that you have previously internalized (passive language), and quickly draw upon them to convey your own message (active language) in your own style.
For this reason, if you want to be able to output naturally it is absolutely mandatory to have achieved at least a solid understanding of the language and have internalized some intuitive model subconsciously. Input is a necessary prerequisite to output. After all, you cannot say things you don’t yourself already understand. As a rule of thumb, if you find yourself constantly flipping through the pages of a dictionary (like jisho.org) looking for a Japanese word to translate a certain English word you have in mind, you’re approaching this incorrectly (more on this later). On the other hand, it is also important to be aware that only relying on input with no practice will not magically make you able to output. You need to practice output to become good at outputting.
I recommend any learner who wants to integrate the output tips on this page to at least have reached the second step of the. Ideally, you should be at the point where you can consume some Japanese media somewhat comfortably (within your area of interest), and regularly do so in your everyday life. If you are at a point where you can read a random Japanese sentence and go “This sentence feels off to me” then you’re ready to output. You don’t need to know what feels off, or why, but you need to have that feeling within you to be able to apply it to a lot of the exercises I will be introducing later.
DISCLAIMER: While some may say you should wait to output until you’re “ready”, I don’t think this matters as much. You should not force yourself to output but there’s also no real downside if you want to engage in conversation (ideally with a native speaker) even at a low level of intuition. However, keep in mind that it will be much easier and way more beneficial if you prioritize your input early on, before you jump into output.
Before we dive into practical advice and exercises to improve on output, it is a good idea to actually understand what output is and which aspects one needs to focus on for it to be “natural”.
As I already mentioned in the introduction, I will not touch on actual pronunciation and phonetics. These are incredibly important aspects of the language if one wants to appear natural, but they belong to an entirely different area. I’m only focusing on the production abilities of a language learner: their ability to put together pieces they acquired from immersion in a manner that appears natural.
With this in mind, we can recognize three fundamental components of natural output:
- Grammatical correctness
- Correct use of words and collocations
- Appropriate social framing and behavior
This is probably the easiest part to understand and one that most people would likely already know about. Simply put, you need to be able to output grammatically correct sentences. To give a simple example, if you say something like 美味しいだ, it would come across as extremely wrong because you cannot put だ after い-adjectives. Any native speaker would be able to instantly point out that this sentence is wrong and it would give them a feeling of discomfort.
Note, however, that sometimes correct output can be grammatically incorrect. For example you might see a native speaker say things like すごい美味しい instead of すごく美味しい. While it’s technically grammatically wrong, it’s actually become normal modern Japanese usage to the point where if you were to say すごく美味しい instead it could come across as weird.
Knowing which part of grammar must be correct and which part is allowed to be more flexible also will come to you with experience and intuition. You will internalize the rules subconsciously with enough input, and often you will not even be able to explain why they work like that, just that they feel better that way.
It goes without saying that even if your grammar is completely perfect and natural, your sentences can come across as extremely unnatural if you use the wrong words. On top of it being unnatural, it can even lead to actual misunderstandings, so it’s important to be very careful on this aspect.
Generally speaking, even before we internalize things as grammar, languages are built upon “chunks” or groups of phrases/expressions that a native speaker can recognize as familiar. There are a lot of phrases even in English where if you were to break them down into their own individual components and words, you might not even be able to define what they actually mean.
To give you an example, how often have you seen the verb “to bode” in English without the word “well” after it? And specifically in a negative nuance like “This does not bode well”. If you were to ask me, I would have no idea what “bode” meant in English, but I am very familiar with the expression “to (not) bode well”. Imagine you’re talking with an English learner and they said something like “The weatherman boded good weather for tomorrow”. While, yes, it’s technically a correct sentence, it feels off. At the very least it feels inappropriate for the level of speech and register we expect them to be using. This is because subconsciously our brains are just very advanced predicting machines: when we talk to someone we are constantly playing a mental game of prediction where we try to stay a couple of steps ahead of the conversation in order to know what to expect from our conversational partner. When these collocation rules are broken, we suddenly feel lost and confused, and the language will be perceived as unnatural.
This is also strengthened when it comes to set phrases and expressions like “to kill two birds with one stone”. If I said “to kill two birds with one rock” it would be wrong, despite the meaning being exactly the same.
This point is much trickier to grasp and often gets confused with the previous one about word choice and expressivity. No language exists in a vacuum. There is always some cultural baggage and shared background that every native speaker is aware of and can use as leverage during a social exchange.
Mannerisms, ways to phrase things, indirectness, politeness levels, things to say and not to say, how to ask certain questions, how to comment on certain things, how to break the ice, how to start and end conversations, etc. These are all incredibly nuanced aspects of Japanese that can often feel very alien or unfamiliar to western audiences. Furthermore, a lot of these points can be very hard to grasp for native speakers as well because they often rely on one’s own personal charisma, sociability, extrovertedness, and a lot of other factors.
However, do not confuse this with the sensationalist talking points of “exoticism” and “us vs them” mentality (Why Japanese people? etc) that certain blogs like to write about when it comes to Japan. Japanese is a language like any other and there’s nothing inherently special or magical to it. It does, however, have its own social rules and etiquette that will be different from your native language, and just like any other language you need to be able to internalize them properly if you want to have natural output.
You might be already familiar with some common mistakes like:
- Answering with ありがとう to a compliment instead of being modest (まだまだ, いえいえ, etc)
- Not using polite language when the situation calls for it (not using ます/です with strangers, etc)
- Overusing pronouns (あなた, etc)
- Using the wrong interjections (“uhmm…” instead of “えと…”)
- When asking for a favor, pre-emptively thanking the other person with ありがとうございます instead of よろしくおねがいします
…and many others.
This is also often called “thinking with your English brain”. There’s nothing ungrammatical or necessarily wrong in it, it’s just… not what a Japanese person would say in that situation.
Now that all the introductory fluff is out of the way, let’s actually get down to the meat of it: how does one improve their output? Can we come up with a series of activities that, if done with enough frequency, will lead to having natural output ability?
While I cannot guarantee with absolute certainty that all of them will work for you, I can say that each of the activities I describe here have helped me improve my ability to communicate in Japanese. I’m still not quite there yet, and it’s a very long trek of continuous gradual improvement, but they are all good starting points that are worth considering.
This is not a specific activity one needs to practice, however it is a central pillar to all the other exercises so it is important to introduce it first.
Practicing output will show you your weaknesses in the language. This is what I call “noticing”. Output practice itself is a self-feeding loop not unlike theand will look something like this:
In my experience I’ve met a lot of very different people when it comes to their “noticing” ability. Some can be natural talents and pick up expressions and word choices almost effortlessly by just doing input. Just like that they’ll be able to use them more or less accurately in output too. Some other people, however, need a bit more of an extra push to be able to figure out certain expressions and phrases that would otherwise go unnoticed. This is where output practice comes into play.
When you spend time practicing output, you make that jump from passive to active knowledge awareness, and that in turn kickstarts the predictive engine in your brain that will make you actively soak up new phrases to use in output via immersion. Try to be consciously aware of this process as you go through each exercise, because it will consistently benefit you a lot more if you do that.
This is probably the most obvious and common exercise that most people would already be familiar with: conversation practice.
As simple as it sounds, just find someone to talk to (text chat is good too!) and just practice. Ideally, you want a native speaker or someone who’s at native level because that would give you the best example of natural conversational language. This said, practicing with other learners is not necessarily a bad thing either.
Past the very initial stages of learning (step 1 of), every person is fundamentally different and language learning itself is not a straight line. Someone at approximately your same level of Japanese might be much better than you at some specific thing, and you might be much better than them at some other specific thing. If you regularly chat together, you can help each other reinforce your weak points.
Just keep in mind that most of your conversation time should be dedicated to native speakers. I’ve personally met many learners who only output around other learners and even after years of doing so their Japanese still feels very unnatural (this might not be the only reason, but I bet it does not help).
A few places you can find people to talk to:
- Language exchange discord servers (like )
- Hire a conversational partner/tutor on italki.com
- Language exchange apps like HelloTalk or Tandem
- Online platforms and communities around your hobbies (If you like games, playing online on Japanese servers, etc)
I define “self-corrected writing” as any kind of (usually written) output that you are in charge of correcting yourself.
Self-correcting is an ability that a lot of language learners don’t seem to be aware of having. The reality of it is that getting genuine and useful corrections from native speakers is hard. Most people have no interest in correcting your mistakes as long as what you’re saying is understandable. You may be able to find teachers or tutors that can help you fix certain problems or mistakes if you send them your output compositions, but it’s not a reliable method and it doesn’t “scale” to large volumes over long periods of time (it is also not cheap if you are paying these people to correct you).
What you can do, instead, is take advantage of that internal radar you should have by now that fires “Weird Japanese detected!” alerts every time you read unnatural sentences. Just because you are the person that wrote them, it does not mean that you are not able to recognize that the language feels wrong.
For this activity, you don’t need to be able to know what is wrong and especially how to fix it. All you need to do is just point your internal radar to what you wrote and let it say “this looks good” or “this does not look good”. This is when your noticing engine will become able to pick up better phrases later during immersion. Do this for a long enough time and your ability to output will improve without requiring external assistance from native speakers.
As already mentioned, this entire exercise relies on you having a solid grasp of what natural Japanese looks like. If you have not trained that internal model yet, you will not be able to take advantage of it here.
My personal tip to make this activity worthwhile is to delay your corrections to a later time. Borrowing from the world of art (especially that of music production), we often become able to notice our mistakes only after we let them rest for a while before going back to them with a fresh mind. First, write some Japanese. Then, let it sit there for an hour or two. Go back to it again and try to answer the following questions:
- Do you understand what the sentence is saying?
- If not, do you at least understand what the sentence is trying to say?
- Do you recognize all the words used in the sentence? Do they feel natural?
- Is the flow of the passage logically consistent?
For this purpose, I usually try to get myself to write some Japanese in the morning (like a blog or personal diary, scrap notes, literally anything you feel like writing about) and then in the evening before going to sleep I go over them and see if anything stands out.
I like to write them on a google doc page because I can easily highlight the parts that feel off. Here is an example from some notes I wrote almost a year ago (terrible Japanese and all):
Notice that I don’t try to correct (what I think are) my mistakes. I just mark them and move on. There will be a lot of mistakes you will miss. A lot of stuff you won’t realize is wrong. But there will still be plenty that you will notice. That’s where the gradual improvement will happen.
Attentive Listening is a type of specific input activity that focuses on acquiring natural chunks of phrases by listening to audio of native speakers using them in natural conversation.
While during normal immersion you’d just be consuming native content for the pleasure of doing so (watching anime, livestreams, youtube videos, etc) and nothing more, with attentive listening you should be consistently on the hunt for “cool” phrases and expressions that you want to make yours.
For this type of activity, some learning methods talk about having a language parent, and if you have one it’s great, but it’s not required. However, I do recommend sticking to natural unscripted language that is conversation-driven. This means, rather than doing this with anime or movies, try using livestreams or podcasts where there’s at least two people having a natural-sounding conversation. Twitch (or similar) streams that focus on gameplay might be less optimal than categories like “just chatting” videos where the focus is more on the streamer talking to other people (or the audience). Vtubers can also be a good source for this type of content (if you’re into that, my personal recommendation is the monday morning talk videos from Ookami Mio like this one).
For normal immersion I usually recommend content that you are interested in and enjoy listening to. However, since with attentive listening you should be primarily paying attention to the form and the flow of the conversation rather than its contents, it’s okay to choose things that are boring or not as engaging to you as long as their type of language is similar to what you want to mimic yourself.
What I like to do is watch videos (ideally pre-recorded rather than live, because you can pause and rewind), and pay attention to how the speaker moves from one topic to the next. How each sentence is connected to the whole. What kind of back and forth (backchanneling, aizuchi, etc) happens between two speakers. I pay attention to things like jumps and interruptions in the conversation. What does the speaker say when it happens? How does he say it?
OPTIONAL: Using a tool like sharex I like to record a few seconds of interesting audio for every exchange I want to acquire and then I build anki cards with only the audio (no sentence, no front, no back, etc, just the audio file). Every day I go through all those anki cards and re-listen to what I recorded. Once I feel like I am familiar enough with that phrasing, I just delete the card. Over long periods of time, I notice I start using similar phrases myself in my output. It’s also okay to record similar (or the same) expressions used in slightly different contexts.
It’s good to keep in mind that this kind of activity is relatively focus-heavy as you need to be constantly paying attention and be on the hunt for such chunks of language. For this reason it’s hard to keep it up for long sessions, I recommend maybe 15 minutes a day (just my personal vibes). If you notice your mind drifting to other stuff, getting bored, or just falling back to normal listening, just stop and do something else. Don’t force it.
Predictive listening is the natural evolution of Attentive Listening. The premise is almost entirely the same, and so is the content you should be drawing inspiration from (conversational videos, etc).
The only difference is that, unlike with Attentive Listening, we are not actually looking for chunks or expressions, we’re actually trying to predict how the conversation will continue before it happens. For this reason you should be doing this on recorded videos (rather than live streams). Every time the speaker asks a question (either to the audience or someone they are interviewing, etc) pause the video and imagine yourself in that situation. How would you answer that question? I don’t mean just the contents of the question, but even the mannerisms you’d imagine yourself using. Do you handwave? Nod your head? Follow up with a はい? End with a ということで? etc
Then, once you’ve done those 10 seconds of introspection, unpause the video and notice how the other person actually answers it. See if it matches your prediction, or if they do something you didn’t think of doing. What is their facial expression? What is their reaction? Try to take in every small facet of the conversation that is not just the contents of the answer itself.
A fundamental aspect to be able to move words you learned through immersion (passive vocab) to your everyday language (active vocab) is to be able to gauge the emotional response that they evoke in your conversational partner.
Myself personally, there’s many words I learned by reading books that I don’t feel comfortable outputting in a conversation because I’m not yet sure what kind of reaction I will get from someone listening to them. I noticed very often when the time comes that I need to use certain words that I’ve never used before, I hesitate and my sentences break down and in turn become even harder to understand for no reason other than me being uncertain about those words.
Because of this, I think there is some real benefit in trying to consciously use certain words for the sole purpose of testing the water even if we might already know those words aren’t natural in that context.
For this kind of activity I prefer a spoken conversation rather than chat, and especially with a native speaker (you can’t trust other learner’s opinions when it comes to this). Just practice with a language partner or a tutor, and try to express an idea specifically because you want to use a certain word you learned. See what their reaction is. Do they ask you to repeat the word? Do they act surprised? Do they perhaps rephrase what you just said using a different word? These are all very important signals you need to look for.
I’ve struggled with my inability to do this for the longest of time and I feel like I’ve only recently started to feel like I’m able to keep up with this in Japanese like I did when I first learned English, so this kind of advice can still be somewhat hard to properly internalize and understand depending on what kind of learner you are and how far you are in your journey.
Furthermore, a huge chunk of the population apparently does not have an internal monologue at all, so if you are in that group (I am not), then this tip will probably not apply to you and you can move to the next point.
This said, the basic idea for this so-called “internal narration” activity is to consciously shift your inner monologue to be 100% in Japanese as you go through your everyday life. This is great to get you aware of certain words and phrases you don’t know how to use. Try to describe what you’re doing in real time in your head as you do it. Think “I will now brush my teeth” as you go brush your teeth, “I am now walking the dog”, “I am making some tea and then I will play some videogames”, “I wonder what I can make for dinner tonight”, etc.
This activity can be “dangerous” if you try to force it too early when your internal model is still not quite ready, because it can make you internalize certain phrases that are literally translated from English (or whatever is your native language). However, once you’re advanced enough that you already started outputting and putting into practice some of the activities I mentioned before, building yourself an internal monologue can be very useful. It can help build up your own Japanese personality and language identity which is fundamental to become able to output naturally (more on this later).
I admit, this is a bit weird, and it might be just me, but I’ve decided to write it here anyway because it’s been great advice for myself and I’m sure it will benefit others too (also reading is great).
There’s many well-researched benefits when it comes to reading as you’re about to fall asleep and most of them have to do with a bunch of stuff I’m not familiar with (brainwaves, etc) so I won’t get too much into those details. This said, I have noticed another huge benefit in reading Japanese books right as I am about to go to sleep, preferably on real paper or an ebook reader, not on your phone. More often than not, as I drift asleep after having read a book, I am much more aware of how my “Japanese brain” sounds. In my mind I go over various phrases and words, often subconsciously, that I have just read in the book. Sometimes when I’m very into a story (and also very tired) I notice that my brain starts dreaming that I’m still reading, and my internal monologue shifts to Japanese naturally as I fill in the gaps of whatever story I was reading at the time.
A feature of the language that often gets overlooked is the feeling of belonging to a certain group or community. After years of taking conversational lessons with a Japanese tutor, I noticed that while it gave me a good base to stand on, it didn’t make my output any more natural or spontaneous. Going through rote exercises or just surface level conversation is simply not enough to acquire good natural output (at least it wasn’t for me).
A lot of defining features of what makes a language feel natural come from subculture and situational usage that is appropriate only within certain groups of peers you identify with.
We are much more apt at picking up expressions and jargon when we feel like we belong to a group. We are part of a community, we have friends, people we respect and want to imitate, and we also want to be accepted by them. All of these aspects are necessary to make the language actually yours and create your own Japanese personality. Without this, it is almost impossible to transition into natural language output. See this amazing example from Stephen Krashen on the topic. __
Just as what you’d normally do to find conversation practice occasions, try to find a place where you can “belong”. It’s probably one of the hardest tips to put into practical use as everyone is different, but it is also one of the most important things one should be aware of because just doing this will almost entirely overshadow every other point of advice I’ve given in this page.
Specifically, try to make your output activity not a function of output itself, but rather find a goal that is beyond just wanting to practice the “language”. Find people you want to talk to because you want to discuss whatever topic interests you. You want to discuss the latest anime episode you just watched, you want to coordinate a raid in an MMO, you are trying to find an easter egg in a newly-released videogame, etc. Do all of that in Japanese, but not because it’s in Japanese. Do it because from now on that is going to become your new Japanese identity and personality. Transitioning your hobbies to 100% Japanese is not only just a function of input. It needs to become a function of your output as well.