On が – Tokyo accent adds a slight nasal sound when a が-line character comes in the middle of a word. Something like “nga”. It’s annoying to those of us who are just trying to learn the language, but people aren’t going to run screaming if you don’t do it. They’ll still understand you.

On えい – it represents a long-E sound most of the time, but there are words where it’s pronounced as written; E-I. For example 丁寧 (ていねい) = “teinei”, not “teenee”.

On silent characters – characters are never silent, but you can often find う-sounds and し being crushed to near-silence or truncated, for example です spoken as “des” or the すけ male name ending as “ske” or おかし (candy) spoken as “okash”.

I wouldn’t. I don’t, either. But then, I’m not a linguist. =)

The people who do the audio for Core 6000 seem to lean more towards “nga”; since I’ve done so many of those flash cards, that’s the way I’ve come to say it too. It’s not really that difficult to do – just think of saying words like “ring” and “sing” – it’s like you’re holding that final “ng” sound, and then saying “a”. Also, even though the Core 6k people generally say it more like that, they seem use the other pronunciation for some words, so really, it shouldn’t matter what way you say it.

“On えい – it represents a long-E sound most of the time, but there are words where it’s pronounced as written; E-I. For example 丁寧 (ていねい) = “teinei”, not “teenee”.” – Not sure how correct that is. When I’ve heard the word, as far as I can remember it’s just been with standard long vowels. Some people seem to pronounce long vowels with hints of the “extension kana” coming in at the end.

In the end, pronunciation is most easily figured out just by listening to natives speak. If they pronounce it some way, then that’s a way it can be pronounced. If you hear lots of people pronounce it a different way, then maybe that is a more common/”standard” pronunciation. One thing I’ll say though, if you’re comparing the pronunciation between Koichi and actual native speakers, side with the natives – while he obviously has more experience than you in the language, he’s learned/learning it just like you, so he’s not going to be perfect in every respect, if you get me.

Fair enough – it’s not like I’ve heard *everyone* pronounce it, anyway :P  At the end of the day, people pronounce things differently to other people, just the way it is. Just think of the breadth of accents in the UK (or US if you’re that side of the pond). Imagine a Japanese native learning English, and trying to judge which is the correct pronunciation by comparing people from Scotland, England, Wales and Ireland :D

“Does the ‘ng’ pattern only apply to ‘ga’, or does it apply to the whole ‘ga’ column?”
Y’know, that’s something I was mulling over when I was writing that last reply haha. Just trying to think of examples. Thinking about the example sentence audio that goes with Core 6k: きぎょう and げき/しげき/げきじょう get pronounced “ki-ngyou” and “ngeki/shi-ngeki/ngeki-jou”, whereas ぐうぜん gets pronounced “guuzen” with the hard g. Bear in mind the audio for Core 6k is recorded by at least 3 different people, though the examples still show it can vary, that there’s no set way to pronounce the ‘g’ sound. I get that you want to pronounce it “correctly” sooner rather than later, but it’s one of these things that you just get an ear for and you pick up.

Yeah, Core 6k is an Anki deck. It’s a list of the 6000 most frequently used words (can’t remember where they got the data for that). There was a flashcard site a while ago called smart.fm that was the SRS of choice for TextFugu when it first started – which was switched to Anki once smart.fm became a paid service called iKnow – and that’s where Core 6k originated, but some kind folks spent lots of time and effort converting it to a free deck on Anki. Should be almost top of the list if you have a look for it in Anki’s shared decks section, since it’s so popular. It was smart.fm’s main attraction back in the day, I think. It used to be split into 2 parts: Core 2000 and Core 6000, but now I think they changed it to Core 1000, Core 2000, etc, so if you’re looking for it, it might be under different names.

…You’re right. Just went and had a look through my deck, and sure enough, げき has a ‘hard g’ while えんげき has the ‘nasal g’. Granted it was different speakers, but on comparing it with every other word that contains 劇, the pattern holds (every word containing 劇 being：劇、演劇、劇場、悲劇、喜劇). Why didn’t I notice that before…? Guess I’ve been saying it wrong all this time then XD Still, like you said, that’s the Tokyo accent. Others may be similar or may vary (most likely).

Everyone just listen to Joel from now on; he knows far too much :P