リスニングパワーには基本的な問題があり、それは英語の発音を音素ベースで教えていると言う事です。基本的な間違いです。
ペリー氏は音声から母音を切り離し、その音を学習者と比較することで発音を正しくできると考えております。それがリスニングパワー診断ソフトの基本です。
英語の音声には音素が規則正しく並んでおりません。音声学の音素とはそのような音があると言う概念の音をベースにしております。
もちろん音声には音の境界がありませんから、切り離す事は無理です。仮に切り離せたとしても言語音は相対的な音ですから学習者は厳密には全部違う母音を持っています。
そのような音をソフトで診断したり、判断する事はできません。それを無理にやると発音が苦しくなります。
そのために、言語を含む英語の音声認識は音素ベースではありません。科学的にはフォルマントの最初の3つの遷移が大きな役割を果たしていると考えられています。
すると、発音で大事な事は音素を正しく調音する事でなく、音声のフォルマントをどう変化させるかなのです。そのような変化を教えるソフトは現存しません。その理由はスコット氏のように音をダイナミックに捉えるのでなく、静的な音で捉える方が容易だからです。
スコット氏の母音を分離したと言うのは、言語でも比較的安定している母音の部分を切り出していると思われます。しかし、言語音の認識で大事な部分はその安定した音でなく、フォルマントの変化している部分であり、音の変局点が重要です。
関係のない母音を切り出し、それを比較して発音が正しとか、間違いとのは、言語音を理解していない証拠です。
スコット氏は発音の専門家と言われておりますが、残念ながら音声認識の基本さえも分かっておりません。そのような人が発音を教える事はできません。
彼には次のようなメールを送りました。返事が来たら公開します。
・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・・
October 13, 2011
Dear Mr. Scott Perry
My name is Keizo Sakurai who is teaching English on sound stream base.
After I checked your wesite, I found several mistakes, however, today I would like to forcus only on speech sound aspect in which you are speciliazed as an instructer.
I am sure that your teaching method is based on phonemes or basic speech sound units, because you claimed that you isolated vowels to teach pronounciation.
I hope, as a linguistics specialist, you know a following book, Linguistics, SIXTH EDITION (An Introduction to Language and Communication). In this book, speech sound is defined as continuous streams of sound.
Steven Pinker of Havard wrote in his book that phonemes do not exist in physcal manner but just illusion. This means English natives do not speak on phoneme base. The speech sound is not a serial of phoneme as you think.This also means you can not isolete nor prove to be right phonemes, since a phoneme does not have unique properties as you try to do so.
In short, phoneme-base learning is not a scientific approach. I hope you do know that no English native learn pronounciation based on phonemes as you are curretly teaching.
People just can not speak English on phoneme base. An average speaker can pronounce at a speed of about 25 to 35 phonemes per second; it is impossible to articulate 25 to 35 phonemes in one second.
I am teaching English pronouciation on sound stream base with amazing results. The reason why all your studetes, including Mr. Yokota, have a little strange pronouciation because, I am sure, you taught them on phoneme base.
I studied your speech proof training software in depth. In your video, you said you isoleted the vowel sounds. How possiblly could you do that? Speech sound is just a continous change of a stream as I explained earlier, how did you know where to separate all vowels?
When we look into human speech sound, it has a linearity and the segmentation problem. In a spectrogram of the phrase "I owe you", there are no clearly distinguishable boundaries between speech sounds. If there are no boundaries then you cannot isolete vowels or phonemes. You are simply cutting a sound out the way you like without any physical consistency.
Although listeners perceive speech as a stream of discrete units, this linearity is difficult to be seen in the physical speech signal. Speech sounds do not strictly follow one another, rather, they overlap. A speech sound is influenced by the ones that precede and the ones that follow. This influence can even be exerted at a distance of two or more segments.
Having disputed the linearity of the speech signal, the problem of segmentation arises: one encounters serious difficulties trying to delimit a stretch of speech signal as belonging to a single perceptual unit. This can be illustrated by the fact that the acoustic properties of the phoneme /d/ will depend on the production of the following vowel because of coarticulation.This means that you just cannot isolate vowels or consonants.
The most difficult part of pronounciation is the coarticulation, and my sound stream based training will solve all these problems that many of Japanese students might have.
I want to remid you that coarticulations occure not only between words but also occures between all sounds.
In youe speech proof training software, you said “I know the freqency of sound because statistic of the sound, aproximately 110Hz”. Does this mean that each sound has unique specific physical propertis?
Are you saying that a tall and fat guy has identical vowls with those of a skinny little girl?
I am possitive that a tall and fat guy has away lower volws than a skinny little girl.If this is the case, how can you calim your software is scientific and accurate.
I am afraid that you only tentatively isolated your own personal vowles which are only applied for yourself. What I mean is that each person has own vowels. That means a specific vowel does not have unique properties or all vowels are relative. If I wish to speak good English, then I should use or find my vowels not yours. I even do not hav a choice which vowls I could use, rather my physical conditions will make itself like everybody else.
Please tell me how your software can prove student’s pronouciation right. If you cannot isolate vowels then you just cannot define vowels prove to ber right.
In your website, you are explaining that “Listening power TOEIC uses echo technology, rapid repetition, multilevel sound recognition, and subliminal stimulation to improve your listening skills.”
I have never heard of the word “multilevel sound recognition” and how your Listening Power helps students to improve their listening skill.Please be a little more specific about that multilevel sound recognition.
Human speech recognition is not based on static phonemes, but based on the dinamic change of sound stream.
Many scientists agree that first three formants are the most important cue to speech identity for vowels and some consonants (such as stops); most important element is not specific frequency of formants but formant transitions.
Different individuals produce the same sound in different ways. Because of this, different phoneme categories overlap. We need to interpret speech in reference to the talker. This is call talker normalization, thus we do not have to be exact on each sound when we speak. It is not necessary to prove each sound we speak, like you do it in your speech proof training software.
I will wait for your explanations of speech perception, and your justificatiion of selling your speech proof training software to improve listening skill.
In your website you claimed that “there is one proven teacher and one proven method that works”, then you have to prove your speech proof training software can really prove the right pronouciation. Please explain how could you claime that your method is proven. As far as I know,you have not proved anything yet.
It is a time for you to prove your software is accurate and scientific, but do not tell me who you have tought in the past any more.
Please let me know if you find any errors in my findings, I am willing to correct my websites and a blog as soon as you prove I am wrong.
Sincerely yours,
Keizo Sakurai