I have gotten this question several times lately from my dad. "So what is it exactly that you do all day long?" Well, aside from going for a run, checking email, facebook, etc. obsessively, and spending way too much time on this blog, I do actually do schoolwork!
I've spent the last several weeks chopping up sound files. After I got it down to what I wanted, I then went through and labeled all the words and all the individual sounds that I'm measuring. It looks like this:
Times a ton of words for each person, times twenty people, times three different samples for ten of those people.
And then you see all those red dotted lines? Those are formant frequencies.
Now I'm in the process of finding a script that will automatically measure where they are 1/3 and 2/3 of the way through the vowels. (I used to measure them by hand, one by one by one. It would take me to the end of the summer if I did that again!)
And what does this have to do with linguistics? Well, then I can take those measurements and make pretty charts of people's vowels, like this one from a pilot study:
And that will be an objective way to measure how people change their speech when they're imitating these other accents. So in the above chart, we have 6 different people with their normal accents and also imitating a Southern accent. It illustrates the fact that the Southern vowel /ay/ has a shorter glide and a significantly lower F1 onset (it also has significantly longer duration). In layman's terms, when people say the word "I" with the Southern accent, it sounds more like "ahhh" than the diphthong "a+i".
And in a nutshell, daddy-o, that is what I do all day :)