Yes! I think your comment is related to what we call "stress shift".
Quoted above is the report I posted a week ago. However, I could not get many people involved (though I started also asking people on Facebook communities) last week. It clearly shows how nice and warm the LibriVox friends have been to my research. Starting from this week, some publishers will be involved in this research, so I am hoping that I can keep up the pace of gathering participants.
-> Yes, it happens to me, too. As a native speaker of Japanese, I try to talk to international students in Japanese so that they can practice their second language. Most of them understand what I am saying during the lectures, but when they come to my office, their listening ability goes down tremendously. Later on I realized that it was ME who is causing the unintelligibility of the talk; I unconsiously speak differently when I have a friendly conversation in my office. I think this unconsious difference is hard for non-native speakers.MichaelMaggs wrote: ↑June 4th, 2020, 6:19 am I realised as I went through that I'm not at all consistent in my pronunciation/stress of apparently similar words.
-> What a nice idea! This will also reduce the number of participants' finger movements. It's only some centimeters of moving the mouse from the "Record" icon to the "Stop" icon, but it means a lot when it's accumulated to 359 words. I will pass your suggestion to the technical support. Thanks for your kind advice.MichaelMaggs wrote: ↑June 4th, 2020, 6:19 am You could reduce that by removing that button entirely, and instead having a single start/stop toggle button.
A few of the words I had to smile at, in particular, "mischeivous," which is a personal pet peeve of mine, because everyone I know pronounces it wrong. 
Ooh, that's interesting! I agree there's only one correct way of saying it, though I'm not sure whether that's the same as yours.mightyfelix wrote: ↑June 5th, 2020, 1:39 pm A few of the words I had to smile at, in particular, "mischeivous," which is a personal pet peeve of mine, because everyone I know pronounces it wrong.
because the UK younger generation (in 1999) favoured #2 over #3, while the overall population in the UK still favoured #3 over #2 at that time.MichaelMaggs wrote: ↑June 5th, 2020, 2:26 pm In the UK I mostly hear 1 and 2, so probably 3 is not that usual in British English.
Of course 1 is right!MichaelMaggs wrote: ↑June 5th, 2020, 2:26 pmOoh, that's interesting! I agree there's only one correct way of saying it, though I'm not sure whether that's the same as yours.mightyfelix wrote: ↑June 5th, 2020, 1:39 pm A few of the words I had to smile at, in particular, "mischeivous," which is a personal pet peeve of mine, because everyone I know pronounces it wrong.
There are already lots of recordings on https://noriko-nakanishi.com/sounds/index.php, and they seem to fall mostly into three categories:
1. People who say MIS-chiv-us (three syllables)
2. People who say mis-CHEEV-i-us (four syllables)
3. People who say mis-CHEE-vus (three syllables).
In the UK I mostly hear 1 and 2, so probably 3 is not that usual in British English.
Of course, 1 is right.
Almost everyone I know, here in the US, uses 2. Drives me crazy. 
I'm guessing that the age breakdown was not available for the US? It's a shame, because I'd almost put money on #2 being preferred by the younger generation here too. Most of the people that I interact with are now in their thirties or younger, and they're the ones I hear using #2. I don't know whether I really hear the older people I know use the word mischeivous at all!nakanishi wrote: ↑June 5th, 2020, 5:41 pm Dear Devorah and Michael,
Thanks for the exciting discussion!
Back in 1999, the proportion of the people pronouncing "mischeivous" was:
1. MIS-chiv-us (three syllables)
US 67%; UK(all) 65%; UK(younger) 49%
2. mis-CHEEV-i-us (four syllables)
US 11%; UK(all) 15%; UK(younger) 29%
3. mis-CHEE-vus (three syllables)
US 22%; UK(all) 20%; UK(younger) 22%
Yes, I will make sure to mention in my thesis that the accent varieties exist not only among regions and generations, but also within the same person. Researchers in the field of sociolinguistics may call it "register" (in the case of the example I indicated above, the differences between the accents used in lecture talks and in friendly conversations), but there is a lot more to it, which I will categorize later on.commonsparrow3 wrote: ↑June 5th, 2020, 8:48 pm As others have mentioned, there are a few words that I pronounce sometimes one way, sometimes another - (Caribbean, as a notable example) -
For the current survey, I asked participants to record only one way to pronounce each word, and sometimes it must be hard to decide which one, especially when different pronunciations are used interchangeably. I am hoping that this interchangeability will be represented by the distributions of the people who chose to pronounce a certain word in a certain way. In other words, rather than asking participants to ask "Out of ten times, how many times would you pronounce this word this way?", I will gather enough data and say "Out of 100 recordings, this word was pronounced this way XX times, and this way YY times".commonsparrow3 wrote: ↑June 5th, 2020, 8:48 pm so I just randomly picked one way or the other, trying to decide which was a bit more likely.
.Yes, you are right. The US data is based on a survey conducted by a Japanese phonetician in collaboration with the UK survey team back then, but unfortunately, there is no publication about her great research. And I am starting to understand how hard it was for her to finish gathering all the data across the US, and then publish the results, especially because English is her second language (like myself).mightyfelix wrote: ↑June 5th, 2020, 9:32 pm I'm guessing that the age breakdown was not available for the US?