I was watching a reality-type TV show where a woman spoke Ebonics. Out of curiosity, I replayed portions of her speaking with closed captioning enabled to see how her speech would be interpreted.
Two questions came to me.
I assume subtitling/closed captioning is done automatically via voice-to-text technology? If this is correct, how did this woman’s pronunciation of the word “ask” (which she clearly pronounced “ax”) get transcribed correctly and repeatedly in the subtitle to “ask”?
Also, she used the phrase “I was fittin’ to eat lunch”. At least, that’s what my ears heard, and “fittin’” is how I’ve always interpreted that phrase. But the subtitle displayed it as “fiddin’”, which doesn’t make sense to me.
“Fittin’” I’ve assumed is a variation of “fitting” (not far from “fixin’”). But what could “fiddin’” relate to?
This one is particularly puzzling because I again came across a similar phrase, on a different show, and this one also displayed the word “fiddin’”.
mmm
I’ve imagined a scene in a comedy movie where a voice-activated device can’t interpret African-American accents and dialect correctly, causing one frustrated user to complain about the “muf***ing honky computer!”
I doubt your first assumption is correct. My guess is that they have a human doing the transcription. As a minimum, they have a human monitoring the output and making corrections as needed. If this wasn’t the case, you would see a whole lot of errors (like you occasionally will see on live broadcast CC). For a reality show, my guess is they would take extra steps to make sure the transcription was accurate in order not to offend.
I had a beautiful response written, then the effing page just blanked, and all was lost.
Captioning is much better than it used to be. Live TV can have problems. I have my favorite channels which are more reliable in captioning. I have also had situations where there is something I really WANT to see, and the captions are spotty, screwed up, or missing entirely. That’s when I turn off the TV and walk away.
Youtube captioning is dismal, and I don’t bother. At all.
~VOW
Just FYI “ebonics” isn’t really the preferred term anymore. Typically this type of language you’re describing is known as AAVE, African-American Vernacular English.
Variations on the same expression, ‘fixing to’, also written as finna, fidna, fixna and I’m sure many other ways. It’s dialect so there’s not really a ‘prescribed’ spelling.
It’s true.
Several years ago, the company I worked for had a “retreat” for middle-management types, complete with all-inclusive resort and paid-for activities. One of the activities was skeet-shooting (trying to shoot clay targets, or clay pigeons, out of the air with a shotgun). Well, the launcher for the targets was voice-activated (made in France, the range manager said, which may, or may not, be important). The range manager demonstrated by shouting “BIRD!” and ka-chung, a clay pigeon was launched across the shooting area. It would also respond to “PULL!”, as IME that is the more traditional signal for launching.
Well, there were about 8 of us in the group, with one Scotsman, the rest were Americans. This was in Texas, and while the Americans might not have all been native Texans, I suspect everyone in the group (other than the Scot) had lived there for at least a decade or more. Anyway, the voice-activated launcher worked flawlessly, until the Scot got up there. Having a natural Scot accent, he would pronounce “bird” more like “baird”, so perhaps that had something to do with it, but the thing just wouldn’t work for him. We had him try “PULL!” a few times (more like “pool” for him) but that wouldn’t work, either.
Maybe the elevator was of French manufacture as well.