The Japanese surely have some weird game shows and this voice recognition challenge is something else. Each contestant has to sound out the word that’s selected from the box in English, which is easy enough for those of us that are born and raised in the USA, hopefully, but isn’t quite the case for those to whom English is not a common language. Only one of them gets every word right, and you can tell from the way she pronounces them that she’s had courses in English and is quite fluent in the language. One might almost say she was a ringer since she’s obviously bilingual and can speak Japanese just as well if not better than English. The game is kind of odd but it does present a challenge that’s unique to the Japanese people, especially those that do not speak English well or at all.
If you’ve ever listened a native Japanese individual that is not well-versed in English or still has a heavy accent then you will notice that they have an issue with reciting r’s and l’s in their speech. It’s not a defect at all but instead just a part of the accepted speech patterns that has to do with how they form their sounds and their letters when speaking them. This is something that has been passed down throughout generations and has become a norm within the country. It makes communication and understanding their English a little difficult at times but once you get used to the sounds and the manner in which they speak it becomes less of an issue and more a matter of just listening.
The voice recognition challenge though is kind of funny since what they’re saying and what the recognition software picks it up as is highly amusing. Words like rhythm, curry, and rock and roll become something completely different when the contestants don’t pronounce the r sounds the same way that they’re spoken in English. In Japan the r and the l are basically the same sound which is why it sounds so odd to American ears. We tend to differentiate between the l’s and r’s by forming them differently as we speak, so they become two very different letters and sounds as they’re spoken aloud. In Japan however they tend to make both l and r sound very alike in their speech and for voice recognition software that’s attempting to break their speech down into English it’s very telling when it comes to how they speak.
A few of them however seemed like they spoke rather clearly and the recognition software still didn’t recognize it. This leads me to believe that while the software is very much on the money it still isn’t perfect since it can’t allow for variables in a person’s speech patterns that might come close enough to the desired word. Of course the software doesn’t really work in that manner since it has no idea what a person is going to say before they say it. It’s all just best guess until it’s given the word.
Tell us what's wrong with this post? How could we improve it? :)
Let us improve this post!