Here you can hear the correct pronunciation: 待つ

Look in this thread how to listen to the pronunciation of other words.

stplush wrote:

So I always thought it would be read as matsu, but I was reading Tae Kim and he has it romaji as matu.

One more reason to just avoid romaji.  Tsu and Tu are both つ.

Except that people are more likely at first to use the new Japanese sounds they learn when they are written in kana (instead of defaulting to their native pronunciation of the alphabet combinations). Not unlike pronouncing English with katakana sounds.

For pronunciation, listen.  Listen without understanding.  Listen without paying attention.  But, listen. 

That's gonna matter much, much more than character set.  I loathe ローマ字, but honestly, there's no pronunciation magic in 仮名.

By "the new Japanese sounds they learn", I meant aural input.  Students are more likely to get the right pronunciation when they associate this aural input with kana from the start. The sound has a habit of morphing into something different when they learn with romaji.  I understood your point and was adding that kana are especially beneficial at the early stages as most students do get aural pronunciation input. Those who already pronounce Japanese don't have as much of a problem with different styles of romaji. (Similar to fluent people having less problem with all kana writing.) This is why tu vs tsu presents more of a difficulty to learners. Sorry if I wasn't clear.

Unfortunately ローマ字 seems necessary for L1 people when originally mapping the 仮名 -- unless they do Remembering the Kana with audio cues only. ;p -- Except it's useful to know the parallels for typing, at any rate, but I'd rather keep that input aspect as specifically functional and subliminal a kinetic memory dynamic as possible.

So after that, I'd try to avoid ローマ字 for learning Japanese words/readings, to keep the alphabet subordinate to the Japanese writing system/audio for the acquisition of phonetic segmentation skills of morae/syllables (going by Coulmas/Olson, et al., writing system is crucial for their development for reading comprehension [and listening, but I hope that's implied]), and also to help avoid processing each alphabet letter independently as the brain is wont to do (Pelli/Farell/Moore, Dehaene), and summoning those well-mapped grapheme-to-phoneme relations from your alphabetic language. Too many unjustifiable fail points, to para-borrow a popular RevTK phrase. ;p Also, some research from Besner & Hildebrandt on kana has me interested in the possibilities of phonographic iconicity for L2 learners whose L1 is an alphabet, but I'll save that for a rainy day...

No idea if I just disagreed or agreed with any other posts in this thread, hehe.

Last edited by nest0r (2009 December 10, 8:21 pm)

jonjimbo2000 wrote:

I have noticed my high school students writing things like Tiba and not Chiba and using tu instead of tsu in place names and people's names. 

What romaji system are they taught at elementary school?

Nihon-Shiki (the japanese system). It makes more sense, grammatically.

yudantaiteki wrote:

Just to be clear, I'm not saying that beginners should use romaji, I just think the notion that romaji leads to bad pronunciation (and kana does not) is wrong.

writing system is crucial for their development for reading comprehension [and listening, but I hope that's implied]

I doubt it, the idea that a writing system is crucial for the development of listening comprehension would not find much support in most mainstream articles.

Actually, there's plenty on the topic. I really kind of loathe it when you invoke 'the mainstream' so vaguely and flippantly to dismiss what others say. Admittedly, you might catch me getting too excited and selective about various studies and it's a kind of running joke with me that everything I say is rather arbitrary/speculative/futurist, but you won't find me dismissing others and making vague absolutist statements, so much as expressing what my intuition/logic tells me in conjunction with critiques of references and/or opposing references.

Particularly relevant, studies have shown that Japanese children's awareness transitions from syllables to morae when acquiring kana, and this helps them shape their metalinguistic awareness/phonetic segmentation skills. This occurs with literate individuals in every literate culture, speech develops alongside literacy. Second-language learners likewise use the writing system to develop their reading in conjunction with aural input and vice versa in a bidirectional process.

I think it's clearly possible to use ローマ字 to learn Japanese, but it makes it harder than it needs to be, as you've already developed strong phonemic links with the letters of the alphabet and won't be using the alphabet very much when reading Japanese. Obviously you can offset this, perhaps methods you're familiar with have a heavier emphasis on speech in the early stages, but I think this leads to imbalances in terms of orthographic/phonetic/semantic mapping that must be corrected later, when they could have been avoided. Thus my above recommendation.


The Effect of Kana Literacy Acquisition on the Speech Segmentation Unit Used by Japanese Young Children
Phonological Analysis Abilities of Chinese and Japanese Children
Kana literacy acquisition and speech segmentation units
Can orthography influence second language syllabic segmentation?: Japanese epenthetic vowels and French consonantal clusters
Orthography shapes the perception of speech
Orthographic influences in spoken word recognition
Orthography, vision, and phonemic awareness
What Writing Represents
Reading spoken words: Orthographic effects in auditory priming
Use orthography in L2 auditory word learning
The remarkable inefficiency of word recognition (Nature)
The Reading Brain
Word Recognition Depends on Script
The Relationship between Phonemic Awareness and Cue Weighting in Speech Perception
Orthographic input and second language phonology
Effects of hanyu pinyin on pronunciation in learners of Chinese as a Foreign Language


Semi-related: http://forum.koohii.com/viewtopic.php?pid=80909

My point being, when you've developed a system of phonemic awareness with the alphabet, with various lexical access strategies centering around parallel letter processing, fairly direct grapheme to phoneme conversion rather than morae, then when you're using text to develop your reading and listening, you'd do best to just stick with kana and encode the orthographic-semantic/orthographic-phonetic/phonetic-semantic relationships that way. Clearly, ローマ字 is used often in Japan and I think it has its uses, but as things stand I'm simply saying that I hope one tries to use the native writing system as much as possible, as early as possible... Hope it doesn't sound like I'm always dissing on the alphabet. English is my one true love.

Last edited by nest0r (2009 December 10, 11:51 pm)

nest0r wrote:

Unfortunately ローマ字 seems necessary for L1 people when originally mapping the 仮名 -- unless they do Remembering the Kana with audio cues only. ;p -- Except it's useful to know the parallels for typing, at any rate, but I'd rather keep that input aspect as specifically functional and subliminal a kinetic memory dynamic as possible.

One could go through the Core2000 w/ dictation. That's how I learned hiragana. I then learned katakana over the course of a few months watching random tv.

I can't agree with you. Kanji is a very nice writing system. It may work better for Chinese than for Japanese, but for Japanese it works much better than an alphabet. I have not talked much with Japanese people about it, but my Taiwanese friends regard learning the kanjis as something very easy and very enjoyable to learn. And my pain with Kanji, after RTK, was never to identify or remember how to write the characters, but the new readings that never cease to appear.

yudantaiteki wrote:

nestor: Are you saying that illiterate speakers of a language cannot listen/hear their native language as well as literate ones?  The fact that the Japanese writing system is the way it is is an accident of history, not a carefully designed approach.

Some of those articles you linked were about reading, not listening, and the ones that were about listening didn't seem to be saying that someone hears better based on a certain script.  For instance, Japanese children having a different speech segmentation unit after reading kana is not necessarily saying that they hear *better* -- I'm not sure that being able to segment Japanese into mora means that they're hearing it better, just that they're able to connect the spoken language to the writing system they're learning.  Phonemic segmentation is important for reading (at least in a transition stage to reading in chunks), but in listening, the spoken language is too fast and processing takes place in real time -- for a second language learner it may help, but I don't think first language learners need it.

I've seen the reverse as well, people who don't know that ん has several different pronunciations in Japanese because they were too hung up on "one kana = one mora".

IceCream wrote:

kanji is so much better than any other writing system because of its visual cues, so its not a choice i would ever think of...

To me the Japanese writing system is one of the worst, if not the worst, systems currently in use in the world; even if it has some advantages, the extreme complexity and huge amount of work it takes to learn it aren't worth it.

You've too much of an 'oral fixation'. ^_^ The alphabet only seems better if you look at it phonetically and in the short-term, but kanji's complexity allows for synergistic superiority. I do think its teaching could be improved from what I've heard, as it doesn't require advanced tech to implement methods similar to what we've used, but I fear its development is being hampered by certain perspectives...........

As for the rest...

Speech perception is impacted by knowledge of the orthography. In other words, the ability to differentiate words is refined, by modulating the continuous input of speech for lexical access through the more skillfull perception of word boundaries. That is to say, the phonetic awareness that writing enhances, as it has since its invention, not only improves reading comprehension but listening comprehension and pronunciation.

My point being, again, that if you've developed these particular and different orthographic-phonetic frameworks in the mind, one ought to just use that writing system enough to transition to the basic units of the new one, minimize interference/fail points.

That's if you're using the writing system, rather than avoiding it because kanji and kana are soooooooo harddd!! 8'(

You can take or recommend the long route if you like, though.

Last edited by nest0r (2009 December 11, 12:44 pm)

yudantaiteki wrote:

So you would agree with the statement that an illiterate person listening to someone talk will have worse comprehension than a literate person?

Kanji are hard to learn.  I don't think you can deny that.  The fact that this forum and RTK itself exists are proof of that.  I personally enjoy learning kanji, but I do not enjoy seeing motivated, intelligent students struggle constantly with the writing system, some of them to the point where they give up altogether.

I don't think the positive benefits of literacy on speech perception are a clearly quantifiable inverse of the negative impact of illiteracy on speech perception, too many variables.

What I am saying is that if you're going to become literate in a language that's new to you, and you acknowledge that literacy refines speech perception, you ought to use it as soon as possible to take advantage of it, and especially do away with the original writing system if you can and esp. if said writing system represents a rather different speech system than what you're used to.

I don't think kanji are hard--hard-er than individual letters, yes. I think the site exists because people want to learn kanji, not because it's intrinsically difficult.

Last edited by nest0r (2009 December 11, 1:04 pm)