9 ms·

It's worth bearing in mind that the source is Japanese. All input methods for the Japanese language are a compromise. Inputting kanji on a physical keyboard is

It's worth bearing in mind that the source is Japanese. All input methods for the Japanese language are a compromise. Inputting kanji on a physical keyboard is nowhere near as fluid as inputting Latin characters - you're constantly toggling between inputting kana and selecting the kanji options presented by the IME. The prediction and correction technologies of smartphone keyboards are a very good match for kanji and hànzì input. As a second-language user of Chinese, I find it considerably faster and easier to input hànzì on a smartphone than with a physical keyboard. The context switching between inputting pīnyīn and selecting hànzì is much less expensive when the hànzì are presented directly above the on-screen keyboard. The prediction and correction algorithms seem to be far more intelligent on mobile, which largely compensates for the slower and more error-prone tactile experience. It is my understanding that most young Japanese people prefer the flick input method, which is a refinement of the old keitai input method used on featurephones with numeric keypads; they are often startlingly quick at using this method, but it poses a far higher switching cost when moving to a QWERTY-derived physical keyboard. I find it entirely plausible that the flick method could simply be inherently superior. https://www.youtube.com/watch?v=8V-za9LT_30 https://www.youtube.com/watch?v=8V-za9LT_30

natch 8y ago
On a tangent here but curious, if you know: what's the closest input method you can get to flick for Chinese? I realize the languages are quite different (might be botching this, but my fuzzy understanding for Japanese is it has kanji+katakana+kana+romaji, where Chinese has only hanzi + pinyin|zhuyin/etc.) Do you think flick is compatible with Chinese?
- emodendroket 8y ago
  They don't have that. They do have handwriting recognition.
  - natch 8y ago
    Well the question was what do they have that comes closest. They do have a lot more than just pinyin and handwriting. If you don't know, it's fine to just say that, or stay away from the discussion, especially when you seem to be answering with such certainty on something where you didn't read the question.
    - emodendroket 8y ago
      They have nothing that resembles the tenkey input because they have nothing that resembles kana, unless you want to count bopomofo. If you want to make some correction, rather than just condescending to me, feel free.
      - natch 8y ago
        Sorry I just find statements of the form “nothing exists” a bit irksome especially when the person talking doesn’t know what exists as can be seen by looking at the matter at hand. Chinese has plenty of ten key input systems, some alive, some dead, since the mid 1990s at least for phones and possibly before that for the keypads accompanying computers. It’s been a very fertile space for innovation and I’d be extremely surprised to meet anyone who has kept up with it, so forgive me for being skeptical of your claim of perfect knowledge of a negative. I was asking a different person who sounded like he knew something, and you answered. I did mention zhuyin which is commonly understood informal shorthand for zhuyinfuhao, the more formal word for bopomofo. There are component based systems as well that work with radicals, strokes, quadrants of the characters, and all kinds of zany stuff. That being said, I am not very familiar with Japanese and didn’t know about the the flick method.
rntksi 8y ago
Reading your comment really put the news article in proper context. Thanks.
laurieg 8y ago
This is definitely a big factor. When you look at a Japanese computer keyboard [1] you might see lots of Japanese characters on the keys. Almost nobody actually uses these to type. Instead, if you want a か(ka) character then you type k,a on the familiar qwerty keys. These phonetic characters then get converted into the more complex Chinese-origin characters as you go along, allowing you to disambiguate homophones. The flick keyboard removes the need for typing 2 Roman letters to make one Japanese letter, instead you just have a single flick. Almost all young people use the flick keyboard and I definitely think it is faster. Cast your mind back to when you first encountered a computer keyboard. I remember hunting for seconds to find letters in this unfamiliar arrangement. This is where many Japanese young people are. There was never a computer in their house and now they are heavy mobile users. The qwerty keyboard is not everyday for a lot of people. As to general computer literacy among Japanese teens: I teach a first year general English course at a Japanese university. The students are drawn from all different faculties so I feel it is a pretty good informal sample. I tried to get students to do an online survey by putting a web address on a slide. Over half do not know what a web address is and draw no distinction between search bar and address bar. [1] https://en.wikipedia.org/wiki/Japanese_input_methods#/media/File:Surface_type_cover_JIS_keyboard_layout_blue.jpg https://en.wikipedia.org/wiki/Japanese_input_methods#/media/...
- mwcampbell 8y ago
  > Cast your mind back to when you first encountered a computer keyboard. I remember hunting for seconds to find letters in this unfamiliar arrangement. How early did you learn to type? As an American born in 1980, I learned in first grade, and I don't remember it being a struggle. So maybe it's really important to learn early. FWIW, I'm visually impaired, so looking at the keyboard wasn't a practical option.
  - ghaff 8y ago
    Conversely I was born about 20 years earlier and I didn’t touch a keyboard until I was a senior in high school.
  - laurieg 8y ago
Noumenon72 8y ago
That's fascinating. Is the video representative of texting speed, or is this the fastest human in the world? Are they making choices that would take effort in a different system, or just clicking the next suggested letter?
- ericdykstra 8y ago
  The video is using simple example sentences where the prediction is always correct, so actual typing means choosing the correct kanji at intervals rather than typing the entire sentence at once, but otherwise it's pretty representative of what a normal person can do.
- jdietrich 8y ago
  >Is the video representative of texting speed, or is this the fastest human in the world? From what I understand, he's on the faster end of the normal range. >Are they making choices that would take effort in a different system, or just clicking the next suggested letter? With Japanese kanji or Chinese hànzì, there's no practical way to directly input such a large range of characters. Users type a phonetic spelling, then the input method editor presents them with a menu of characters with a corresponding pronunciation. Chinese mostly uses a system of phonetic transliteration based on the Latin alphabet (pinyin), whereas Japanese speakers use both a Latin-based system (romaji) and a native Japanese system of syllabic characters (kana). The flick method shown in the video uses directional gestures to input kana. For example, if I'm trying to type the Chinese word for bread (面包), I'll input the word as it is pronounced, "mianbao". On mobile devices, a list of predicted characters will appear above my keyboard; on a computer, a numbered list will appear beside my cursor. I select the characters I was intending to input by tapping on mobile, or by pressing the corresponding number key or clicking on a computer. The choice of characters invariably requires some amount of human input, because there are many homophones (different words with the same pronunciation). This method of text input can often be quite slow and cumbersome, so good prediction and correction algorithms are crucial. The input method is constantly guessing which characters you want; if it's not aware of context, it'll make bad guesses and require a lot more manual selection and correction. Good input method software can predict entire phrases and is very resilient to typos.
astatine 8y ago
This is insightful. I see behaviour which princely has similar roots. In India, I have seen people who never input any Indic language text on a regular keyboard do that many times on a phone soft keypad. The prediction and multiple methods to input, including phonetic Latin, makes it significantly easier than the older keyboard based methods. Besides there were several non- standard keyboard methods which didn't help!
AndrewKemendo 8y ago
Seems like there's room then for a touch based input system to augment/replace the standard keyboard for these users.
zenir_ 8y ago
Sounds like finally a good usage for the mac touchbar
- vslira 8y ago
  My thoughts exactly. Maybe Apple was a couple of steps ahead and added an amazing feature for the ever growing Chinese market. Who knows?
- ben-schaaf 8y ago
  Feels like it would work better with the Zenbook Pro's touchpad, seeing as it's about the size of a Smartphone.
- kalleboo 8y ago
  The touchbar does support this (for Japanese at least) but I still can't get used to looking down from the screen onto the keyboard and still use the standard method
ericdykstra 8y ago
I can attest that flick-method typing is fantastic for Japanese. I'm over 100 WPM typing in English, so using standard romaji-phonetic input on a keyboard is still the fastest for me, but flick-typing is bar-none the best option for a touch-screen. You don't even have to bother with using the modifier button (for the ゛or ゜markers) because the prediction will guess what you meant. It's much easier than even typing my native English as far as phone input goes.
innocenat 8y ago
That's why it's so sad this[0] 2016 Google April's Fool stays an April's Fool 0: https://www.google.co.jp/ime/furikku/ https://www.google.co.jp/ime/furikku/
- mey 8y ago
  It sounds like even just the google flick keyboard on a tablet as an input device to a computer would be useful. Have a Qwerty keyboard center, a flick keyboard left and mouse right? (Or personal preference of course)
- delroth 8y ago
  What do you mean, "stays an April's Fool"? There are circuit schematics as well as software for it, all available under an open source license: https://github.com/google/mozc-devices/tree/master/mozc-furikku https://github.com/google/mozc-devices/tree/master/mozc-furi...
  - innocenat 8y ago
    There's no commercial product? I'd love to be proven wrong though.
    - DennisAleynikov 8y ago
      It's a kit :)
- kps 8y ago
  Seems like a DataHand would work well for Japanese.
Kagerjay 8y ago
Great insightful comment, I've never really typed on anything but an IME keyboard on a computer, so I wasn't familiar with how easy it is to type kanji / hanzi on a phone IME keyboards (for Japanese) has always confused me, since sometimes if you input hiragana or katakana it would translate to the kanji equivalent because japanese is kind of weird in that they have 3 written types. Thats just a quirk with Japanese itself though. I'm am glad that English doesn't have such an issue though. Then you have Romaji, the English - Japanese phonetic equivalent Its also important to note when doing these comparisons, Japanese has a significantly lower information informational density than that of other languages. It also has a low reading speed rate too (informational rate), but very high speech rate (e.g. how fast you can talk with the language). English is actually very high across the board. https://www.quora.com/Whats-the-most-efficient-highest-information-transfer-rate-spoken-language-If-you-took-a-corpus-of-text-translated-into-a-bunch-of-different-languages-which-version-would-take-the-least-time-to-read-aloud https://www.quora.com/Whats-the-most-efficient-highest-infor... The article here has an image of the table I am thinking of (its the first response) for the data, from lyons et al. But, it does omit certain languages that are of interest as well, namely arabic. The rate at which you can write arabic and achieve the same level of information by hand is very closely related to that of English shorthand https://en.wikipedia.org/wiki/Shorthand https://en.wikipedia.org/wiki/Shorthand. This might not tie directly to texting, but its worth mentioning because notetaking efficiently requires quickly capturing information. Some people prefer by hand others by typing etc I don't really know all the slang terminology in Japanese all that well, but in English we have things like brb lol btw lmfao roflcopter fam. I don't really know what the Japanese equivalent at all of this is, and I wander whether if you were to compare Japanese Slang to English slang / Ebonics, which input type method is superior on a phone? The japanese flick input method or a traditional qwerty keyboard. My bet is on the latter though (QWERTY) Like I know instead of saying わたしは you could just shorten it to ぼくは (only if your male though). Which means "I am ..... {{doing something}}". You could just say instead "{{doing something}}" too, which gets the same point across. In mandarin you would use 我是 which is pinyin is typed "woshi" which is the equivalent.
carindistory 8y ago
As a Japanese native speaker who has to enter many Japanese texts on a daily basis both on a PC and on a smartphone, I feel obliged to point out that the quality of an IME (a piece of software providing possible kanji/alphabet sequences for a given phonetic relization), is far better on a PC (its built-in MS-IME, to be precise) than on a smartphone (iOS). I started looking for some example to show how bad iOS's IME is, and I found it for the first try: it returns a wrong candidate for the first suggestion for "かんじをにゅうりょくする" (to enter kanji), returning "感じを入力する" (to enter sense) rather than the correct "漢字を入力する". Note that "漢字" (kanji) and "感じ" (sense) have the exact same phonetic relization: Both of them are pronounced as kanji. It seems as if iOS's IME does not take into account any contexts at all. If it did, how could it have calculated entering sense (?) is more likely than entering kanji? This kind of absurd error would rarely happen with Microsoft's IME and it always stresses me out when entering long texts in Japanese on a smartphone.
- kalleboo 8y ago
  The iOS IME definitely uses context but I think the default training just isn't very good. If I try your example it starts suggesting 感じ, but once I add を it changes to 漢字を. But this is also why there is a large market in third party keyboards/IMEs, even on Windows. From the classic ATOK to the modern Google Japanese IME
- TazeTSchnitzel 8y ago
  I think iOS's IME expects you to segment things yourself, rather than typing several words strung together and then pressing the suggest button?
- hrktb 8y ago
  I think this might be the training (or lack of) of your device. I get it right on mine: https://imgur.com/a/83VCSwo https://imgur.com/a/83VCSwo Bear in mind I use a mac wih the same icloud account, and have years of data on it. Part of it must be shared. I haven’t use MS’IME in a long time, but I remember it being only marginally better than Apple’s. The main differentiators for me where names (places, stations, peoples).
polygot 8y ago
I'm on a bit of a tangent, but it appears that a lot of people are beginning to use voice-recognition to recognize Japanese and Chinese, or are sending the actual voice clips instead of manual input methods such as handwriting recognition. Granted, this is my very limited experience. Perhaps this could be a competing way to help communicate in foreign languages.
rahimnathwani 8y ago
Which IMEs have you tried with a physical keyboard? I don't type much Chinese on a keyboard these days but I recall that the Sogou IME for Windows had far more accurate ranking of suggestions than did the built-in one.
fireattack 8y ago
>I find it considerably faster and easier to input hànzì on a smartphone than with a physical keyboard As a native Chinese speaker I find the opposite. >when the hànzì are presented directly above the on-screen keyboard It's basically the same for most of "proper" PC IME. They start to appear when you're typing pinyin, and then you choose with numbers or space (for first one). Showcase: https://i.imgur.com/gQGKw11.gif https://i.imgur.com/gQGKw11.gif So, IMHO there is no obvious advantage that smartphone's IME has over PC's here. And the speed of typing on physical keyboards beats smartphone by a mile, overally speaking. --- However, I DO found that most of Japanese (which I'm a second-language user) IME on PC I tried having the problem that, you have to press some key (normally `enter` or `space`) to start "convert", and even another input to start choose between candidates, which is very tedious because you have to keep pressing `space` which IMO shouldn't be necessary (Note: It was how most Chinese IME (Like Zhineng ABC) worked 10 years ago, but they got rid of redundant inputs later.) Showcase (MS JP input, I knew it's not the best, so feel free to let me know how other IME function in these scenarios!): https://i.imgur.com/hwml1Sf.gif https://i.imgur.com/hwml1Sf.gif Notice that I have to press space once first to enter "convert mode" (which breaks down your inputs to groups), and then press space again (for the first group) to make the candidates appear. I really don't get why it can't be like the Chinese IME.
- cthalupa 8y ago
  >you have to press some key (normally `enter` or `space`) to start "convert", and even another input to start choose between candidates This doesn't match my experience, and I just double checked on Windows 10 and OSX 10.13 - I do not believe this experience has changed in at least the past 7-8 years. I switch to Japanese hiragana input, and type 'sake', and 酒 appears as soon as I hit space or escape. I do not hit space or escape or anything prior to typing 'sake'. If I type 'yoroshikuonegaishimasu'　and hit space/escape/enter it becomes 宜しくお願いします, properly converting hirigana to kanji where it should. I am using the default IMEs that come with Windows and OSX.
  - fireattack 8y ago
    I believe I was not clear, but your description is exactly what I said. * You type in "sake" * Press space once, it becomes "酒" * If you want to choose OTHER candidates (such as "鮭"), you have to press space again to show the numbered candidate list. Please see this showcase: https://i.imgur.com/rXxjMRs.gif https://i.imgur.com/rXxjMRs.gif >I do not hit space or escape or anything prior to typing 'sake' I don't mean you need to hit anything prior to typing Romajis. But you need to hit it twice afterwards to get the numbered candidate list. To be fair, on Win10's default IME, it does provide a predicted suggestion list before you press any spaces, which is nice; but for some reason this list is different from the formal candidate list, that it is not numbered and you have to hitting tabs repeatedly to choose from them:
_bxg1 8y ago
This makes a whole lot more sense. I couldn't imagine how a student could get through high school and all its essays without once thinking a physical keyboard might be an easier option.
cthalupa 8y ago
I'm unsure how relevant my experience is as someone who has learned (some) 日本語 as a second language, but I find that I do not have any more trouble typing in Japanese on the keyboard than I do English. The IME in Windows and OSX is quite good at determining which kanji makes sense in context. In that case, it's as simple as pressing space/escape/enter to confirm the selection. I've always remapped capslock to escape because I'm a vi user, so I tend to gravitate towards that, but there's definitely choice there. I do find the flick method to be significantly faster on a touchscreen, however, I'm still quicker with a regular qwerty keyboard. I'm curious if we would see similar results if swype-style keyboards had as much domination among young people in countries that use Roman characters as the flick method does in Japan.

natch 8y ago

On a tangent here but curious, if you know: what's the closest input method you can get to flick for Chinese? I realize the languages are quite different (might be botching this, but my fuzzy understanding for Japanese is it has kanji+katakana+kana+romaji, where Chinese has only hanzi + pinyin|zhuyin/etc.) Do you think flick is compatible with Chinese?

emodendroket 8y ago
They don't have that. They do have handwriting recognition.
- natch 8y ago
  Well the question was what do they have that comes closest. They do have a lot more than just pinyin and handwriting. If you don't know, it's fine to just say that, or stay away from the discussion, especially when you seem to be answering with such certainty on something where you didn't read the question.
  - emodendroket 8y ago
    They have nothing that resembles the tenkey input because they have nothing that resembles kana, unless you want to count bopomofo. If you want to make some correction, rather than just condescending to me, feel free.
    - natch 8y ago
      Sorry I just find statements of the form “nothing exists” a bit irksome especially when the person talking doesn’t know what exists as can be seen by looking at the matter at hand. Chinese has plenty of ten key input systems, some alive, some dead, since the mid 1990s at least for phones and possibly before that for the keypads accompanying computers. It’s been a very fertile space for innovation and I’d be extremely surprised to meet anyone who has kept up with it, so forgive me for being skeptical of your claim of perfect knowledge of a negative. I was asking a different person who sounded like he knew something, and you answered. I did mention zhuyin which is commonly understood informal shorthand for zhuyinfuhao, the more formal word for bopomofo. There are component based systems as well that work with radicals, strokes, quadrants of the characters, and all kinds of zany stuff. That being said, I am not very familiar with Japanese and didn’t know about the the flick method.

rntksi 8y ago

Reading your comment really put the news article in proper context. Thanks.

laurieg 8y ago

This is definitely a big factor. When you look at a Japanese computer keyboard [1] you might see lots of Japanese characters on the keys. Almost nobody actually uses these to type. Instead, if you want a か(ka) character then you type k,a on the familiar qwerty keys. These phonetic characters then get converted into the more complex Chinese-origin characters as you go along, allowing you to disambiguate homophones. The flick keyboard removes the need for typing 2 Roman letters to make one Japanese letter, instead you just have a single flick. Almost all young people use the flick keyboard and I definitely think it is faster. Cast your mind back to when you first encountered a computer keyboard. I remember hunting for seconds to find letters in this unfamiliar arrangement. This is where many Japanese young people are. There was never a computer in their house and now they are heavy mobile users. The qwerty keyboard is not everyday for a lot of people. As to general computer literacy among Japanese teens: I teach a first year general English course at a Japanese university. The students are drawn from all different faculties so I feel it is a pretty good informal sample. I tried to get students to do an online survey by putting a web address on a slide. Over half do not know what a web address is and draw no distinction between search bar and address bar. [1] https://en.wikipedia.org/wiki/Japanese_input_methods#/media/File:Surface_type_cover_JIS_keyboard_layout_blue.jpg https://en.wikipedia.org/wiki/Japanese_input_methods#/media/...

mwcampbell 8y ago
> Cast your mind back to when you first encountered a computer keyboard. I remember hunting for seconds to find letters in this unfamiliar arrangement. How early did you learn to type? As an American born in 1980, I learned in first grade, and I don't remember it being a struggle. So maybe it's really important to learn early. FWIW, I'm visually impaired, so looking at the keyboard wasn't a practical option.
- ghaff 8y ago
  Conversely I was born about 20 years earlier and I didn’t touch a keyboard until I was a senior in high school.
- laurieg 8y ago

Noumenon72 8y ago

That's fascinating. Is the video representative of texting speed, or is this the fastest human in the world? Are they making choices that would take effort in a different system, or just clicking the next suggested letter?

ericdykstra 8y ago
The video is using simple example sentences where the prediction is always correct, so actual typing means choosing the correct kanji at intervals rather than typing the entire sentence at once, but otherwise it's pretty representative of what a normal person can do.
jdietrich 8y ago
>Is the video representative of texting speed, or is this the fastest human in the world? From what I understand, he's on the faster end of the normal range. >Are they making choices that would take effort in a different system, or just clicking the next suggested letter? With Japanese kanji or Chinese hànzì, there's no practical way to directly input such a large range of characters. Users type a phonetic spelling, then the input method editor presents them with a menu of characters with a corresponding pronunciation. Chinese mostly uses a system of phonetic transliteration based on the Latin alphabet (pinyin), whereas Japanese speakers use both a Latin-based system (romaji) and a native Japanese system of syllabic characters (kana). The flick method shown in the video uses directional gestures to input kana. For example, if I'm trying to type the Chinese word for bread (面包), I'll input the word as it is pronounced, "mianbao". On mobile devices, a list of predicted characters will appear above my keyboard; on a computer, a numbered list will appear beside my cursor. I select the characters I was intending to input by tapping on mobile, or by pressing the corresponding number key or clicking on a computer. The choice of characters invariably requires some amount of human input, because there are many homophones (different words with the same pronunciation). This method of text input can often be quite slow and cumbersome, so good prediction and correction algorithms are crucial. The input method is constantly guessing which characters you want; if it's not aware of context, it'll make bad guesses and require a lot more manual selection and correction. Good input method software can predict entire phrases and is very resilient to typos.

astatine 8y ago

This is insightful. I see behaviour which princely has similar roots. In India, I have seen people who never input any Indic language text on a regular keyboard do that many times on a phone soft keypad. The prediction and multiple methods to input, including phonetic Latin, makes it significantly easier than the older keyboard based methods. Besides there were several non- standard keyboard methods which didn't help!

AndrewKemendo 8y ago

Seems like there's room then for a touch based input system to augment/replace the standard keyboard for these users.

zenir_ 8y ago

Sounds like finally a good usage for the mac touchbar

vslira 8y ago
My thoughts exactly. Maybe Apple was a couple of steps ahead and added an amazing feature for the ever growing Chinese market. Who knows?
ben-schaaf 8y ago
Feels like it would work better with the Zenbook Pro's touchpad, seeing as it's about the size of a Smartphone.
kalleboo 8y ago
The touchbar does support this (for Japanese at least) but I still can't get used to looking down from the screen onto the keyboard and still use the standard method

ericdykstra 8y ago

I can attest that flick-method typing is fantastic for Japanese. I'm over 100 WPM typing in English, so using standard romaji-phonetic input on a keyboard is still the fastest for me, but flick-typing is bar-none the best option for a touch-screen. You don't even have to bother with using the modifier button (for the ゛or ゜markers) because the prediction will guess what you meant. It's much easier than even typing my native English as far as phone input goes.

innocenat 8y ago

That's why it's so sad this[0] 2016 Google April's Fool stays an April's Fool 0: https://www.google.co.jp/ime/furikku/ https://www.google.co.jp/ime/furikku/

mey 8y ago
It sounds like even just the google flick keyboard on a tablet as an input device to a computer would be useful. Have a Qwerty keyboard center, a flick keyboard left and mouse right? (Or personal preference of course)
delroth 8y ago
What do you mean, "stays an April's Fool"? There are circuit schematics as well as software for it, all available under an open source license: https://github.com/google/mozc-devices/tree/master/mozc-furikku https://github.com/google/mozc-devices/tree/master/mozc-furi...
- innocenat 8y ago
  There's no commercial product? I'd love to be proven wrong though.
  - DennisAleynikov 8y ago
    It's a kit :)
kps 8y ago
Seems like a DataHand would work well for Japanese.

Kagerjay 8y ago

Great insightful comment, I've never really typed on anything but an IME keyboard on a computer, so I wasn't familiar with how easy it is to type kanji / hanzi on a phone IME keyboards (for Japanese) has always confused me, since sometimes if you input hiragana or katakana it would translate to the kanji equivalent because japanese is kind of weird in that they have 3 written types. Thats just a quirk with Japanese itself though. I'm am glad that English doesn't have such an issue though. Then you have Romaji, the English - Japanese phonetic equivalent Its also important to note when doing these comparisons, Japanese has a significantly lower information informational density than that of other languages. It also has a low reading speed rate too (informational rate), but very high speech rate (e.g. how fast you can talk with the language). English is actually very high across the board. https://www.quora.com/Whats-the-most-efficient-highest-information-transfer-rate-spoken-language-If-you-took-a-corpus-of-text-translated-into-a-bunch-of-different-languages-which-version-would-take-the-least-time-to-read-aloud https://www.quora.com/Whats-the-most-efficient-highest-infor... The article here has an image of the table I am thinking of (its the first response) for the data, from lyons et al. But, it does omit certain languages that are of interest as well, namely arabic. The rate at which you can write arabic and achieve the same level of information by hand is very closely related to that of English shorthand https://en.wikipedia.org/wiki/Shorthand https://en.wikipedia.org/wiki/Shorthand. This might not tie directly to texting, but its worth mentioning because notetaking efficiently requires quickly capturing information. Some people prefer by hand others by typing etc I don't really know all the slang terminology in Japanese all that well, but in English we have things like brb lol btw lmfao roflcopter fam. I don't really know what the Japanese equivalent at all of this is, and I wander whether if you were to compare Japanese Slang to English slang / Ebonics, which input type method is superior on a phone? The japanese flick input method or a traditional qwerty keyboard. My bet is on the latter though (QWERTY) Like I know instead of saying わたしは you could just shorten it to ぼくは (only if your male though). Which means "I am ..... {{doing something}}". You could just say instead "{{doing something}}" too, which gets the same point across. In mandarin you would use 我是 which is pinyin is typed "woshi" which is the equivalent.

carindistory 8y ago

As a Japanese native speaker who has to enter many Japanese texts on a daily basis both on a PC and on a smartphone, I feel obliged to point out that the quality of an IME (a piece of software providing possible kanji/alphabet sequences for a given phonetic relization), is far better on a PC (its built-in MS-IME, to be precise) than on a smartphone (iOS). I started looking for some example to show how bad iOS's IME is, and I found it for the first try: it returns a wrong candidate for the first suggestion for "かんじをにゅうりょくする" (to enter kanji), returning "感じを入力する" (to enter sense) rather than the correct "漢字を入力する". Note that "漢字" (kanji) and "感じ" (sense) have the exact same phonetic relization: Both of them are pronounced as kanji. It seems as if iOS's IME does not take into account any contexts at all. If it did, how could it have calculated entering sense (?) is more likely than entering kanji? This kind of absurd error would rarely happen with Microsoft's IME and it always stresses me out when entering long texts in Japanese on a smartphone.

kalleboo 8y ago
The iOS IME definitely uses context but I think the default training just isn't very good. If I try your example it starts suggesting 感じ, but once I add を it changes to 漢字を. But this is also why there is a large market in third party keyboards/IMEs, even on Windows. From the classic ATOK to the modern Google Japanese IME
TazeTSchnitzel 8y ago
I think iOS's IME expects you to segment things yourself, rather than typing several words strung together and then pressing the suggest button?
hrktb 8y ago
I think this might be the training (or lack of) of your device. I get it right on mine: https://imgur.com/a/83VCSwo https://imgur.com/a/83VCSwo Bear in mind I use a mac wih the same icloud account, and have years of data on it. Part of it must be shared. I haven’t use MS’IME in a long time, but I remember it being only marginally better than Apple’s. The main differentiators for me where names (places, stations, peoples).

polygot 8y ago

I'm on a bit of a tangent, but it appears that a lot of people are beginning to use voice-recognition to recognize Japanese and Chinese, or are sending the actual voice clips instead of manual input methods such as handwriting recognition. Granted, this is my very limited experience. Perhaps this could be a competing way to help communicate in foreign languages.

rahimnathwani 8y ago

Which IMEs have you tried with a physical keyboard? I don't type much Chinese on a keyboard these days but I recall that the Sogou IME for Windows had far more accurate ranking of suggestions than did the built-in one.

fireattack 8y ago

>I find it considerably faster and easier to input hànzì on a smartphone than with a physical keyboard As a native Chinese speaker I find the opposite. >when the hànzì are presented directly above the on-screen keyboard It's basically the same for most of "proper" PC IME. They start to appear when you're typing pinyin, and then you choose with numbers or space (for first one). Showcase: https://i.imgur.com/gQGKw11.gif https://i.imgur.com/gQGKw11.gif So, IMHO there is no obvious advantage that smartphone's IME has over PC's here. And the speed of typing on physical keyboards beats smartphone by a mile, overally speaking. --- However, I DO found that most of Japanese (which I'm a second-language user) IME on PC I tried having the problem that, you have to press some key (normally `enter` or `space`) to start "convert", and even another input to start choose between candidates, which is very tedious because you have to keep pressing `space` which IMO shouldn't be necessary (Note: It was how most Chinese IME (Like Zhineng ABC) worked 10 years ago, but they got rid of redundant inputs later.) Showcase (MS JP input, I knew it's not the best, so feel free to let me know how other IME function in these scenarios!): https://i.imgur.com/hwml1Sf.gif https://i.imgur.com/hwml1Sf.gif Notice that I have to press space once first to enter "convert mode" (which breaks down your inputs to groups), and then press space again (for the first group) to make the candidates appear. I really don't get why it can't be like the Chinese IME.

cthalupa 8y ago
>you have to press some key (normally `enter` or `space`) to start "convert", and even another input to start choose between candidates This doesn't match my experience, and I just double checked on Windows 10 and OSX 10.13 - I do not believe this experience has changed in at least the past 7-8 years. I switch to Japanese hiragana input, and type 'sake', and 酒 appears as soon as I hit space or escape. I do not hit space or escape or anything prior to typing 'sake'. If I type 'yoroshikuonegaishimasu'　and hit space/escape/enter it becomes 宜しくお願いします, properly converting hirigana to kanji where it should. I am using the default IMEs that come with Windows and OSX.
- fireattack 8y ago
  I believe I was not clear, but your description is exactly what I said. * You type in "sake" * Press space once, it becomes "酒" * If you want to choose OTHER candidates (such as "鮭"), you have to press space again to show the numbered candidate list. Please see this showcase: https://i.imgur.com/rXxjMRs.gif https://i.imgur.com/rXxjMRs.gif >I do not hit space or escape or anything prior to typing 'sake' I don't mean you need to hit anything prior to typing Romajis. But you need to hit it twice afterwards to get the numbered candidate list. To be fair, on Win10's default IME, it does provide a predicted suggestion list before you press any spaces, which is nice; but for some reason this list is different from the formal candidate list, that it is not numbered and you have to hitting tabs repeatedly to choose from them:

_bxg1 8y ago

This makes a whole lot more sense. I couldn't imagine how a student could get through high school and all its essays without once thinking a physical keyboard might be an easier option.

cthalupa 8y ago

I'm unsure how relevant my experience is as someone who has learned (some) 日本語 as a second language, but I find that I do not have any more trouble typing in Japanese on the keyboard than I do English. The IME in Windows and OSX is quite good at determining which kanji makes sense in context. In that case, it's as simple as pressing space/escape/enter to confirm the selection. I've always remapped capslock to escape because I'm a vi user, so I tend to gravitate towards that, but there's definitely choice there. I do find the flick method to be significantly faster on a touchscreen, however, I'm still quicker with a regular qwerty keyboard. I'm curious if we would see similar results if swype-style keyboards had as much domination among young people in countries that use Roman characters as the flick method does in Japan.