Oh Hey There

so for my speech pathology class, i told my professor about how i plotted my vowels last year and i offered to make some slides and show-and-tell some things about the so-called “vowel space” to our class. i basically go over a basic methodology of vowel measurement, then present my vowel space with some observations and give some take-home big ideas that might be helpful to budding phoneticians. here are the big ideas:

• Vowel space is tricky.
• Vowel space is more relational than absolute.
• Context of a vowel matters but not too much.
• …except for when it does matter (i.e. phonological language difference).

the above image shows my diphthongs—oy in boy, ow in cow, ai in bye—in the vowel space. left is towards the front of the mouth and up is towards the top of the mouth. diphthongs are “double vowels” so this plot shows the heads, tails and trajectories of my vowels. i have something normally referred to as “canadian raising” where i have appreciably different vowels in the words ice and eyes (green vs red), and to a lesser extent, in kraut and crowd (blue vs orange).

Oh how fun. I’ll have to badger one of the authors tomorrow.

Holy fuck it’s a plot of my vowel formants!

I had to record me saying “please say [word] for me” where [word] was one of 109 words. I then I had to to go through the 109 tokens, and with the help of some PRAAT scripting, highlight and label the vowel steady-state of each token and extract an mp3 of the vowel. This took hours.

But look at this plot! I made it!

One nice thing about the F1/F2 plot is that it corresponds to the traditional vowel space with front vowels on the left and high vowels on the top of the plot. Two 20ms samples of each vowel were taken, one at the “head” and one at the “tail” of the vowel. You need two measurements because diphthongs and deviantly diphthongish monophthongs start and end in different places articulatorily and acoustically. This is what the vectors on the plot show. And as you can see, I’ve got a healthy diphthong in bite.